abstract title of thesis: age, growth and recruitment of … · 2017. 5. 2. · recruitments...
TRANSCRIPT
ABSTRACT
Title of Thesis: AGE, GROWTH AND RECRUITMENT OF
HUDSON RIVER SHORTNOSE STURGEON (ACIPENSER BREVIROSTRUM)
Ryan J. Woodland, Master of Science, 2005 Thesis Directed By: Professor David H. Secor, Department of
Marine, Estuarine & Environmental Science
Shortnose sturgeon (Acipenser brevirostrum), an Endangered Species, has
experienced a several-fold increase in abundance in the Hudson River in recent
decades. Age structure and growth were investigated to evaluate the hypothesis that
improvements in water quality during the late 1970s stimulated population recovery.
Specimens were captured using gill nets bi-monthly 2003 to 2004. Annuli in fin
spine sections were determined to form at an annual rate and yielded age estimates of
5-30 years for sizes 49-105cm Total Length (n=554). Hindcast year-class strengths,
corrected for gill net mesh selectivity and cumulative mortality indicated high
recruitments (28,000-43,000 yearlings) during 1986-1992, which were preceded and
succeeded by c. 5 year-periods of lower recruitment (5,000-15,000 yearlings).
Results indicated that Hudson River shortnose sturgeon abundance increased due to
the formation of several strong year-classes occurring about five years subsequent to
improved water quality in important nursery and forage habitats in the upper Hudson
River estuary.
AGE, GROWTH AND RECRUITMENT OF HUDSON RIVER SHORTNOSE STURGEON (ACIPENSER BREVIROSTRUM)
By
Ryan J. Woodland
Thesis submitted to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfillment
of the requirements for the degree of Master of Science
2005 Advisory Committee: Professor David H. Secor, Chair Assistant Professor Christopher L. Rowe Assistant Professor Robert H. Hilderbrand
© Copyright by Ryan J. Woodland
2005
ii
Dedication
I’d like to dedicate this work to my family for their encouragement, support and
limitless patience.
iii
Acknowledgements
First, I’d like to thank my advisor, Dr. David H. Secor, for his direction and
support during my Masters research. His mentorship helped to make my work a
wonderful and exciting experience. I’d like to extend similar thanks to the other
members of my committee, Drs. Christopher L Rowe and Robert H. Hilderbrand for their
much appreciated advice and valuable feedback.
This work would not have been possible without the knowledge and assistance of
Steven Nack on and about the Hudson River. I’d like to thank Scott McGuire and Bob
Murphy for their help in the field and Jody Callihan for his help, advice and
commiseration in the lab. Thanks to Vincent Mudrak, James Henne, Kent Ware, Roman
Crumpton, Haile Macurdy, Carlos Eschevarria and Jaci Zelco for their assistance in
acquiring samples from the USFWS National Fish Hatchery stock. I’d like to thank Drs.
Frank Chapman, Doug Peterson, and Mark Bain for their interest, encouragement and
their dedication to the shortnose sturgeon. Similarly, Steven Leathery and Jennifer
Jefferies of the Office of Protected Resources; Sheila Eyler of the USFWS; Mark Collins
of the South Carolina DNR; Mark Mattson of Normandeau Assoc.; Bill Andrews of the
NYDEC; and also Vicky Kelly and Dave Fischer of the Institute for Ecological Studies
for their assistance in locating and providing data used in some of the analyses included
in this project.
A final, crucial thank you to the Hudson River Foundation, the University of
Maryland Center for Environmental Science and Chesapeake Biological Laboratory for
funding and supporting my research.
iv
Table of Contents Dedication ........................................................................................................................... ii Acknowledgements............................................................................................................ iii Table of Contents............................................................................................................... iv List of Tables ..................................................................................................................... vi List of Figures ................................................................................................................... vii Introduction..........................................................................................................................1
Hypotheses and Goals................................................................................................... 10 Age Determination .................................................................................................... 11 Growth and Mortality Rate ....................................................................................... 11 Year-class Strength ................................................................................................... 12
Methods..............................................................................................................................13
Field Sampling .............................................................................................................. 13 Laboratory Methods...................................................................................................... 16 Age Precision Tests and Age Validation ...................................................................... 17 Growth .......................................................................................................................... 23
von Bertalanffy Growth Model ................................................................................. 23 Gompertz Growth Model .......................................................................................... 24 General Growth Models ........................................................................................... 25 Mark-recapture ......................................................................................................... 25
Gear Selectivity............................................................................................................. 27 Baranov Model.......................................................................................................... 29 Holt Model ................................................................................................................ 30 Regression Model...................................................................................................... 31
Recruitment................................................................................................................... 32 Fall Juvenile Survey.................................................................................................. 34 Environment: Flow ................................................................................................... 35
Results................................................................................................................................37
Field Sampling .............................................................................................................. 37 Age Precision Tests and Age Validation ...................................................................... 41 Growth .......................................................................................................................... 51 Gear Selectivity............................................................................................................. 55 Recruitment................................................................................................................... 60
Fall Juvenile Survey.................................................................................................. 64 Environment: Flow ................................................................................................... 69
Discussion..........................................................................................................................72
Age Determination........................................................................................................ 73 Growth .......................................................................................................................... 77 Recruitment................................................................................................................... 85
v
Gear Selectivity ......................................................................................................... 85 Mortality ................................................................................................................... 86 Age-structure............................................................................................................. 87
Management Implications............................................................................................. 90 Appendix............................................................................................................................93 References........................................................................................................................107
vi
List of Tables Table 1. Estimated status of shortnose sturgeon in eight river systems. Adapted from the National Marine Fisheries Service Recovery Plan for Shortnose Sturgeon (Acipenser brevirostrum), 1998............................................................................................................. 2 Table 2. Permitted capture and sampling schedule under NMFS Permit to Take Endangered Species #1360. .............................................................................................. 15 Table 3. Tag and meristic data of previously marked shortnose sturgeon recaptured during field sampling (from USFWS sturgeon tagging database). FL0 = Fork Length at tagging, FLR = Fork Length at recapture, TL0 = Total Length at tagging, TLR = Total Length at recapture, Wt0 = weight at tagging, WtR = weight at recapture........................ 26 Table 4. Statistical analyses used to diagnose accuracy of age determinations for hatchery raised sturgeon. .................................................................................................. 45 Table 5. Mean Marginal Increment Ratio values (MIR) by sampling month and season ± 95% Confidence interval. Means with any identical letters are not significantly different at α= 0.05 (Tukey-Kramer protected against experimentwise error inflation)................. 49 Table 6. Growth during time at large for recaptures of previously marked shortnose sturgeon during field sampling. ........................................................................................ 56 Table 7. Catch by age class with mean Total Length (TL), catch per mesh, actual (Cactual) and adjusted (Cadjusted) catches of shortnose sturgeon over the course of field sampling. In the bottom row is estimated mortality, Z, calculated for each mesh, actual, and adjusted values, Z used in calculations, 0.22, given in boldface..................................................... 62 Table 8. Mean catch curve residual values grouped by age-class stanzas. See Figure 21 for catch curve. Means with any identical letters are not significantly different at α = 0.05 (Tukey-Kramer protected against experimentwise error inflation). ......................... 65 Table 9. Correlation of yearling recruitment strength index (RSI) values relative to shortnose sturgeon bycatch from the Hudson River utilities sponsored Juvenile Fall Survey, lagged over yearly increments. ............................................................................ 68 Table 10. Precision in ageing studies of Acipenser and other fish species of varying longevity, reported as average percent error (APE) and coefficient of variation (CV). ... 74 Table 11. Growth parameters from the von Bertalanffy growth function estimated for shortnose sturgeon across their range. Boldface row indicates parameters derived during this study. .......................................................................................................................... 80
vii
List of Figures Figure 1. Shortnose sturgeon bycatch catch-per-unit effort (CPUE) from the Hudson River utilities Fall Juvenile Survey 1985-1999. ................................................................. 3 Figure 2. Time series of spring-fall dissolved oxygen conditions in the Hudson River during the 1970s directly downstream of the Albany, NY area. The dotted line (c. 5 mg L-1) indicates an approximate threshold for hypoxia in sturgeons (Secor & Niklitschek 2001). Data from Leslie et al. 1988. .................................................................................. 5 Figure 3. Lateral view of adult (upper image) and juvenile shortnose sturgeon (adult photo courtesy of J.Jensen, www.fishbase.org). Juvenile image has been expanded to show morphological detail (lower image). ......................................................................... 7 Figure 4. Annual distribution of shortnose sturgeon within the Hudson River estuary. Seasonal habitat for life-stages is depicted with reference to the fresh/brackish interface (dashed lines) and the three primary sampling areas (alternating dashed and dotted line). Figure modified from Bain 1999. ....................................................................................... 9 Figure 5. Posterior lobe from a mounted shortnose sturgeon pectoral spine (transverse section) photographed under reflected light conditions at 450 X magnification. The opaque and translucent zones from a single annulus are indicated. Within the dashed box, a false or ‘supernumary’ annulus is differentiated from the ‘true’ annulus.............. 18 Figure 6. Posterior lobe of a shortnose sturgeon pectoral spine (36 X magnification) with annuli enumerated according to their position relative to the outermost annulus (marginal increment). Annuli are demarcated by the circles and increment width for each annulus is indicated by the brackets. .............................................................................................. 21 Figure 7. Catch per unit effort (CPUE) of Hudson River shortnose sturgeon over the course of the sampling period (2003-04). CPUE for each sampling period by gear type is shown in panel. ................................................................................................................. 38 Figure 8. Total catch (n = 587) of shortnose sturgeon over the course of the sampling period (2003-04) by A) weight (kg) and B) Total Length (mm). ..................................... 39 Figure 9. Distribution of estimated ages from interpretable spines (n = 554). ................ 40 Figure 10. Transverse sections of pectoral spines viewed under reflected light microscopy. Common sources of ageing error are indicated by arrows in each image. A) Spine showing initial stages of resorbtion of calcified material, canalization and proteinaceous deposits. B) Inclusion of secondary rays within the posterior lobes; inclusion can vary from two, one or none within the primary spine. C) Inconsistent annulus widths. D) “Crowding” of annuli near edge of spine. ........................................ 42
viii
Figure 11. Age-bias plot of subsampled spines (n = 55) for Hudson River shortnose sturgeon. Mean second ages are plotted with 95% confidence intervals. The 1:1 ratio line is shown. .................................................................................................................... 44 Figure 12. Analysis of age determinations of hatchery-raised shortnose sturgeon (n = 46). A) Paired differences of estimated minus true ages (residuals) plotted against actual age (abscissa) with positive (young fish, 1-6 years) and negative (older fish, 7-20 years) bias clearly visible. B) Estimated ages (y-axis) plotted against actual ages (abscissa) with a linear regression (least squares) line (solid line) relative to the 1:1 line (dotted line). .. 46 Figure 13. Transverse sections of pectoral spines from four different hatchery-raised shortnose sturgeon. A & B) depict hatchery spines that appear to be laying down consistent increments without noticeable deformation or artificial structures. C & D) depict spines that are deteriorated, possess multiple false structures and large calcareous deposits. ............................................................................................................................ 48 Figure 14. Mean seasonal Marginal Increment Ratios (MIRs) with 95 % C.I. A) Overall Marginal Increment Analysis (MIA, ages 5-26). B) Reduced set MIA (ages 10-17). Seasonal MIR values that have different letters above them are significantly different at α = 0.05 (Tukey-Kramer protected against experimentwise error inflation). ...................... 50 Figure 15. Growth curves calculated for mean Total Length (mm) and mean weight (kg) on age (years) with range of size values indicated by vertical bars and denoted by circles. Filled circles were used in the analysis, open circles were not. A & B) Loge growth function; C & D) Polynomial (quadratic) growth function; E & F); von Bertalanffy growth function; G & H) Gompertz growth function....................................................... 52 Figure 16. Size-at-age data for Hudson River shortnose sturgeon (combined sex data). Length–age data given by open circles (A) with mean length-at-age given by closed circles with fitted von Bertalanffy curve for ages 5-30 (B). Weight–age data given by open circles (C) with mean weight-at-age given by closed circles with fitted von Bertalanffy curve for ages 5-30 (D).................................................................................. 54 Figure 17. Growth trajectories of tagged shortnose sturgeon recaptured during sampling. Mean growth rate during time at large was 11 ± 5mm year-1, with trajectories of growth exceeding the mean falling above the shaded area and those below falling within the shaded area. Maximum and minimum observed growth rates are indicated by the arrows............................................................................................................................................ 57 Figure 18. Mean annual growth (mm yr-1) plotted against length at recapture (mm) for shortnose sturgeon tagged and recaptured from the Hudson River. ................................. 58 Figure 19. Selectivity curves for each mesh (10.2 cm mesh: open circles, 15.2 cm mesh: filled circles, 17.8 cm mesh: filled triangles) overlaid on catch values standardized to 1. Plots A-C are curves generated from the Holt model, plots D-F are curves generated from the Regression model, and plots G-I are generated from the Baranov model. ................. 59
ix
Figure 20. Baranov selectivity curves generated for each of the mesh sizes. Optimal capture length lO is denoted by arrows for each mesh size............................................... 61 Figure 21. A) Natural log transformed adjusted catch (y-axis) plotted against age for Hudson River shortnose sturgeon. Least-squares linear regression was fitted to filled data given by filled circles only. Above each symbol is the associated subsample size. B) Catch curve residuals plotted against age (abscissa) with four year-class stanzas denoted by the dashed boxes. ......................................................................................................... 63 Figure 22. Index of yearly recruitment success based upon actual catch adjusted for gear selectivity, effort and cumulative mortality. Relative cohort strengths (y-axis) are plotted against hatch year (abscissa)............................................................................................. 66 Figure 23. Estimated age 1 cohort abundance for Hudson River shortnose sturgeon, ages 5-30 years.......................................................................................................................... 67 Figure 24. A) Hindcast annual recruitment strengths (open bars) versus shortnose sturgeon CPUE from the Hudson River utilities sponsored Fall Juvenile Trawl Survey. B) Trawl bycatch CPUE lagged by 6 years and overlaid on hindcast recruitment strengths. ........................................................................................................................... 70 Figure 25. Time series of mean monthly flow volume (ft3 minute-1, primary y-axis) collected from Green Island, New York (1946-2002, USGS) with hindcasted recruitment strength (secondary y-axis) overlaid on flow for corresponding years (Blow-up window) of 1974-1997..................................................................................................................... 71 Figure 26. Von Bertalanffy growth parameters FL ∞ (primary y-axis) and k (secondary y-axis) plotted against latitude. ............................................................................................ 81 Figure 27. Predictions of Fork Length-at-age from von Bertalanffy growth models (combined sex models) for three different estuaries across the latitudinal range of shortnose sturgeon. Sources for each curve: 1) Dadswell et al. 1984, 2) current study, 3) Dadswell et al. 1984, 4) Dadswell 1979. .......................................................................... 82
1
Introduction
Throughout their range, shortnose sturgeon (Acipenser brevirostrum)
populations have been negatively affected by anthropogenic changes to their habitats.
Decreased water quality, habitat destruction, blockage of spawning runs and
incidental/intentional harvest (Kynard 1997, NMFS 1998, Collins et al. 2000, Secor
& Niklitschek 2001, Root 2002) have effected reduced abundance or localized
extirpations (e.g. Chesapeake Bay) in some instances (Secor et al. 2002). As a result,
shortnose sturgeon were federally protected range-wide in 1973 pursuant to the
Endangered Species Act (ESA) in the United States and are considered a species of
special concern under the Canadian Species at Risk Act (SARA).
Yet, while many shortnose sturgeon populations number in the hundreds
(Table 1), the Hudson River population may be as high as 55,000 fish (Bain et al.
2000). Due to their life history traits (i.e., periodic strategist sensu Winemiller &
Rose 1992), low rates of overall population growth and recovery are expected for
shortnose sturgeon (Boreman 1997; Gross et al. 2002). In contrast, results from
mark-recapture studies on the Hudson River indicate population growth from c.
10,000 individuals in 1980 to c. 55,000 by 1995 (Bain et al. 2000), a 450% increase
over this 15-year period. Shortnose sturgeon bycatch data from the Fall Juvenile
Survey sponsored by a consortium of Hudson River utilities corroborated a
population increase during this period (Figure 1). An elasticity analysis conducted by
Gross et al. (2002) indicates that such rapid population growth is unlikely to be the
result of enhanced survival rates among sub-adult and adult life stages (i.e., due to
protection from harvest). Rather, rapid population growth in sturgeons can only
Table 1. Estimated status of shortnose sturgeon in eight river systems. Adapted from the National Marine Fisheries Service Recovery Plan for Shortnose Sturgeon (Acipenser brevirostrum), 1998.
2,862
361
14,080
38,024
714
33
7,222
18,000
Population Estimate
1,069 – 4,2262004Altamaha*
326 – 4001993Ogeechee
10,079 – 20,3781981-1984Delaware
26,427 – 55,0721995Hudson
280 – 2,8561976-1978Connecticut
18 – 891989-1990Merrimack
5,046 – 10,7651977-1981Kennebec
12,600 – 23,4001973-1977Saint John
95% Confidence IntervalStudy periodRiver
2,862
361
14,080
38,024
714
33
7,222
18,000
Population Estimate
1,069 – 4,2262004Altamaha*
326 – 4001993Ogeechee
10,079 – 20,3781981-1984Delaware
26,427 – 55,0721995Hudson
280 – 2,8561976-1978Connecticut
18 – 891989-1990Merrimack
5,046 – 10,7651977-1981Kennebec
12,600 – 23,4001973-1977Saint John
95% Confidence IntervalStudy periodRiver
3
0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.519
85
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
CPU
E x
10-2
(cat
ch 1
000
m3
-1)
Year
0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.519
85
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
CPU
E x
10-2
(cat
ch 1
000
m3
-1)
Year Figure 1. Shortnose sturgeon bycatch catch-per-unit effort (CPUE) from the Hudson River utilities Fall Juvenile Survey 1985-1999.
4
occur due to changes in first year survival and the formation of strong annual cohorts
(year-classes). These results lead to two questions: 1) ‘Why did the abundance of
shortnose sturgeon increase dramatically in the Hudson River?’ and 2) ‘Why did this
occur during the period 1980-1995?’
The increase in shortnose sturgeon abundance in the Hudson River was
preceded by improved water quality during the previous decade, as mandated by the
Clean Water Act (CWA) of 1970. Prior to the 1970’s, dissolved oxygen (DO)
concentrations in the 60 km stretch of the Hudson River below Albany (c. km from
river mouth [rkm] = 235) dropped precipitously with the onset of summer (c. 2 mg L-
1), reaching a nadir in the early fall (< 1 mg L-1; Figure 2; Leslie et al. 1988). This
stretch of the river, known as the “Albany Pool” (Boyle 1969), coincides with
approximately 40% of the shortnose sturgeon’s estimated nursery habitat (Dovel et al.
1992, Haley 1999). Thus, the hypoxic Albany Pool may have functioned as a
recruitment bottleneck, rendering much of the summertime nursery habitat unsuitable
for juvenile shortnose sturgeon (Secor and Niklitschek 2001). As a result of the
CWA, stringent standards on industrial and municipal effluent precipitated the return
of normoxia to the Albany region by 1978 (Figure 2).
The Hudson River shortnose sturgeon population is centered within the
species’ historical range, which stretches from the St. John River in New Brunswick
southward to Florida’s St. Johns River (Vladykov & Greeley 1963, Dadswell et al.
1984). Present day distribution mirrors historical patterns in the northern region
while extirpations and unknown population status characterizes many populations in
5
0
2
4
6
8
10
12
Jun Jul Aug Sep Oct NovMonth
DO
(mg
L-1)
197819741970
0
2
4
6
8
10
12
Jun Jul Aug Sep Oct NovMonth
DO
(mg
L-1)
197819741970
Figure 2. Time series of spring-fall dissolved oxygen conditions in the Hudson River during the 1970s directly downstream of the Albany, NY area. The dotted line (c. 5 mg L-1) indicates an approximate threshold for hypoxia in sturgeons (Secor & Niklitschek 2001). Data from Leslie et al. 1988.
6
the Mid-Atlantic and southern portion of the species’ range (Dadswell et al. 1984,
NMFS 1998, Collins et al. 2000).
The anadromous shortnose sturgeon (Figure 3) is the smallest of North
America’s coastal sturgeons, and inhabits large, estuarine systems ranging from full
salinity (> 30 psu) to damlocked non-tidal fresh water (e.g., upper Connecticut River)
(Greeley 1937, Dadswell 1979, Taubert 1980, Dovel et al. 1992, Collins & Smith
1997, Kynard 1997). The largest individual on record is also the oldest: a 143 cm
Total Length (TL) female weighing 23.6 kg was captured from the St. John River in
New Brunswick, Canada having an estimated age of 67 years (Dadswell 1979).
Depending upon latitude (in which maturation occurs later in northern latitudes),
males become sexually mature at 2-11 years, and females at 4-18 years (Dadswell et
al. 1984).
Shortnose sturgeon display a punctuated iteroparous spawning strategy in
which adults of both sexes spawn intermittently (c. every 2-11 years), substantially
curtailing reproductive rates in comparison to annual spawners (Dadswell et al. 1984,
Boreman 1997). The spring spawning event follows an upstream migration and
typically occurs over hard substrate in moderate to swift flow velocities (37-180 cm
sec-1) with water temperatures ranging from 9 to 15 OC (Taubert 1980, Dadswell et al.
1984, Kynard 1997). Initially, embryos (1-8 days post-hatch) orient to benthic
structure followed by a shift in orientation towards the water column that initiates a
diurnal downstream dispersal to juvenile nursery habitats (Richmond & Kynard 1995,
Kynard & Horgan 2002).
7
0 10 20 30 40 50 60 70cm
10 cm0
0 10 20 30 40 50 60 70cm
10 cm0 Figure 3. Lateral view of adult (upper image) and juvenile shortnose sturgeon (adult photo courtesy of J.Jensen, www.fishbase.org). Juvenile image has been expanded to show morphological detail (lower image).
8
In the Hudson River and most other estuaries, young-of-the-year (YOY)
shortnose sturgeon make exclusive use of freshwater habitats before becoming
euryhaline (Niklitschek (2001). Larval and YOY nursery habitat in the Hudson
River extends downstream from the spawning grounds near the Federal Dam in Troy,
NY (rkm 246), and encompasses much of the tidal freshwater portion of the estuary
(Bain 1997). Sub-adults and adults form dense seasonal aggregations in well-defined
areas (Dadswell 1979, Dovel 1992, Buckley & Kynard 1985, Kynard 1997). In the
Hudson River, non-spawning adults and juvenile shortnose tend to over-winter near
the fresh/brackish water interface in the Haverstraw Bay region (rkm 54 – 61; Figure
4; Bain 1997). A second overwintering aggregation occurs near Kingston (rkm 140),
serving as a staging area for ovigerous adults participating in the following spring’s
spawning migration (Bain 1997).
Shortnose sturgeon, like many congenerics., are sensitive to hypoxia and
display negative metabolic and behavioral responses in the presence of low ambient
DO levels (Secor & Gunderson 1998, Secor & Niklitschek 2001, Campbell &
Goodman 2004). A recent study by Campbell and Goodman (2004) found juvenile (≤
134 day-old) shortnose sturgeon to be unusually sensitive to low DO in acute tests (24
h), with 50% mortality occurring at 2.2 mg L-1 and 26oC for 134 day-old fish. In the
same study, lethal concentrations for younger fish occurred at similar or higher
concentrations, ranging from 2.2 mg L-1 (104 day-old) to 2.7 mg L-1 (77 day- old).
These results agreed with prior work by Jenkins et al. (1993) who reported a similar
pattern of sensitivity to hypoxia in juvenile shortnose sturgeon.
9
Bra
ckis
h
B
rack
ish/
Fres
h
fres
h in
terf
ace
Non-spawning
adults
Spawning adults
Eggs & larvae
Juveniles
Spring –
Summer –
Winter –
Albany (rkm 235)
Catskill (rkm 172)
EsopusMeadows (rkm 134)
Atlantic Ocean
Spring sampling area
Summer sampling
area
Fall/Early Spring
sampling area
Haverstraw Bay region
N
Bra
ckis
h
B
rack
ish/
Fres
h
fres
h in
terf
ace
Non-spawning
adults
Spawning adults
Eggs & larvae
Juveniles
Spring –
Summer –
Winter –
Albany (rkm 235)
Catskill (rkm 172)
EsopusMeadows (rkm 134)
Atlantic Ocean
Spring sampling area
Summer sampling
area
Fall/Early Spring
sampling area
Haverstraw Bay region
N
Figure 4. Annual distribution of shortnose sturgeon within the Hudson River estuary. Seasonal habitat for life-stages is depicted with reference to the fresh/brackish interface (dashed lines) and the three primary sampling areas (alternating dashed and dotted line). Figure modified from Bain 1999.
10
Hypotheses and Goals
My thesis is that the rapid population growth by Hudson River shortnose
sturgeon observed during the 1980-1995 period was due to improved water quality in
nursery habitats. To evaluate the link between recovering water quality and
increasing shortnose sturgeon abundance, I retrospectively estimated past year-class
strengths from the age structure of the extant population. I predicted that year-class
strengths adjusted for gear efficiency and cumulative mortality effects would be
higher after 1980 than before 1980. The pattern of incidental captures of shortnose
during the Hudson River Utilities Juvenile Fall Survey (1985-2003) provides
evidence that the 1980s were a pivotal decade in which a large pulse of sturgeon
recruited to the adult population (individuals tend to be susceptible to survey gear at
lengths > 500 mm). These anecdotal data support the hypothesis that one or more
very strong year-classes recruited to the population during the 1980s.
High spring freshet volume associated with seasonal precipitation and climatic
conditions (e.g. rate of snow-melt, water temperature, etc.) has been implicated in
favoring spawning success among anadromous species (Stevens et al. 1987, Secor
2000, Jung & Houde 2003). Therefore, I predicted that spring freshet volume in the
Hudson River would be positively correlated with YOY survival and subsequent
recruitment. This secondary analysis was incorporated to examine the possibility that
strong recruitment events during the 1980s stemmed from favorable flow conditions
rather than improved DO levels.
Direct sampling and ageing of Hudson River shortnose sturgeon supplied
estimates of current growth and mortality, age structure of the standing stock, and the
11
relative abundances of previous year classes. In particular, I used shortnose sturgeon
longevity to provide a retrospective analysis of annual recruitments as far back as the
late 1970s. In order to investigate the current demographics of this population and
accurately hindcast recruitment events, I focused my research on three specific
objectives:
Age Determination
The development of an accurate and precise ageing methodology was a critical
goal of my thesis. Although an age determination technique has not yet been
validated for shortnose sturgeon, ageing of Acipenser sp. is typically carried out by
enumerating concentric growth structures on thin cross-sections of the ossified
anterior spine of a pectoral fin (Cuerrier 1951, Rossiter et al. 1995, Stevenson and
Secor 1999). A comprehensive study conducted by Brennan and Cailliet (1989)
analyzed a suite of calcified structures (i.e., pectoral fin spines, opercles, clavicles,
cleithra, scutes and medial nuchals) collected from white sturgeon (A. transmontanus)
and reported that pectoral spines provided the highest level of inter- and intra-reader
precision among all structures evaluated. The use of pectoral spines is also preferable
to other structures because it has been shown to be non-deleterious to shortnose
sturgeon (Collins and Smith 1996).
Growth and Mortality Rate
I estimated present vital rates of the Hudson River population and contrasted
with mark-recapture estimates and literature values from the Hudson River and other
systems. Intra- and inter-system comparisons of vital rates allowed corroboration of
12
age estimates with previous studies. I was also interested in detecting differences in
observed versus expected mortality patterns as a means of evaluating past year-class
strengths (see below).
Year-class Strength
To accurately hindcast annual recruitment strengths, gear selectivity and
population mortality rate were modeled to adjust the current age structure. Annual
year-class strengths were compared and tested for dominant cohorts, temporal trends,
and correlation with other records of abundance and environmental variables.
13
Methods
Field Sampling
Prior to sampling, protocols for capture and handling shortnose sturgeon were
reviewed and approved by the National Marine Fisheries Service Protected Species
Division (Permit No. 1360-01), NY State Department of Environmental
Conservation, and the Institutional Animal Care and Use Committee of University of
Maryland Center for Environmental Science. Our methods adhered to the proscribed
capture, care and handling protocols for shortnose sturgeon published by NMFS
(NMFS 2000).
I conducted field sampling on the Hudson River from November 2003
through November 2004 on a bimonthly basis. Sampling locations and gear
deployments were chosen to maximize catch based upon the annual distribution of
shortnose sturgeon in the system (Figure 4) (Dovel 1992, Dadswell 1984, Bain 1997).
Fall and late winter sampling (November-2004, March/November-2005) was
conducted at Esopus Meadows Point (rkm 134), targeting an over-wintering
aggregation assumed to consist of primarily pre-spawn adults and juveniles. Spring
sampling (April-2005) was carried out at the spawning grounds near Albany, several
kilometers south of the Federal dam at Troy (rkm 245). Summer sampling (June-
August-2005) locations varied but were concentrated in the Catskill area (rkm 172)
because the species tends to disperse during summer (Bain 1997). Based on a
statistical analysis using a test dataset derived from research conducted on shortnose
in the St. John River estuary, New Brunswick (Dadswell 1979), target sampling goals
were set in terms of overall sample size and monthly sample size (D. Secor, pers.
14
comm.). These sampling goals were established to provide reasonable power in
subsequent analyses (e.g., growth rate, mortality, marginal increment analysis) while
minimizing capture and/or handling stress in accordance with our Endangered
Species Permit (Table 2).
Gill nets used for sampling were constructed of #6 single strand, clear
monofilament rigged with a foam-core float line and lead-core line, measuring 0.91 m
high by 30.5 m in length. We attached concrete anchors and floats to the distal ends
of the lead line and float line respectively as a means of anchorage and retrieval.
Mesh sizes of 10.1, 15.2 and 17.8 cm (stretch) were selected based on previous
research, which showed these meshes captured all sizes shortnose sturgeon beyond 48
cm Fork Length (FL) (Dadswell 1979). Nets were set perpendicular to the river
channel during slack tide and allowed to soak from 3 to 60 minutes, depending on the
rate of capture for a given location and day. We employed a YSI® meter to measure
water temperature, salinity and dissolved oxygen at most sampling sites (the meter
was unavailable during summer sampling). Latitude and longitude were also
recorded at each sampling location using GPS.
Captured sturgeon were carefully extricated from the net and immediately
transferred to floating recovery pens. Sturgeon were weighed to the nearest 10 grams
with a spring scale then measured for FL and Total Length (TL) to the nearest
millimeter. During the spring sampling period, an attempt was made to sex fish based
on external features associated with spawning (Dadswell 1979). Sturgeon were
scanned on their dorsal surface with a Passively Induced Transponder tag reader
(Pocket Reader EX®, Biomark, Inc.) to identify internal tags, and were visually
Table 2. Permitted capture and sampling schedule under NMFS Permit to Take Endangered Species #1360.
-
Warm Springs (GA), Bears Bluff (SC), UF
Gainesville (FL)
Newburg-Kingston (rkm 90-140)
Albany-Troy area (rkm 230-245)
Kingston area (rkm 140)
†Haverstraw area (rkm 50-60)
Region
November 1, 2003 –November 30, 2004
Accidental moralityJuvenile/Adult-34*
600
56
36
176
141
191
Actual capture
November 1, 2003 –November 30, 2004
Age validationJuvenile/Adult70
740
June 15 –November 30, 2004
Population age structure, marginal increment analysis
Juvenile/Adult100
May 10 –June 30, 2004
Population age structure, marginal increment analysis
Juvenile/Adult190
March 1 –April 30, 2004
Population age structure, growth rate
Juvenile/Adult190
November 1 –December 20, 2003
Population age structure, growth rate
Juvenile/Adult190
Sampling schedulePurposeLife stageAllowable capture
-
Warm Springs (GA), Bears Bluff (SC), UF
Gainesville (FL)
Newburg-Kingston (rkm 90-140)
Albany-Troy area (rkm 230-245)
Kingston area (rkm 140)
†Haverstraw area (rkm 50-60)
Region
November 1, 2003 –November 30, 2004
Accidental moralityJuvenile/Adult-34*
600
56
36
176
141
191
Actual capture
November 1, 2003 –November 30, 2004
Age validationJuvenile/Adult70
740
June 15 –November 30, 2004
Population age structure, marginal increment analysis
Juvenile/Adult100
May 10 –June 30, 2004
Population age structure, marginal increment analysis
Juvenile/Adult190
March 1 –April 30, 2004
Population age structure, growth rate
Juvenile/Adult190
November 1 –December 20, 2003
Population age structure, growth rate
Juvenile/Adult190
Sampling schedulePurposeLife stageAllowable capture
*No mortality observed, all fish released alive in apparent good condition, †Kingston area sampled due to logistic constraints.
16
examined for external tags. Past and ongoing studies of shortnose sturgeon by other
scientists have applied both external and internal tags to track migration and habitat
use patterns, and measure abundance (Dovel 1992, Bain 1997, NMFS 1998).
A one-centimeter section of pectoral spine was removed from one of the
pectoral fins approximately 1 cm distal from the point of articulation with the pectoral
girdle. This minimized the risk of losing proximal annuli associated with growth
during the early life period while preventing damage to the artery occurring at the
articulation of the spine (Rien & Beamesderfer 1994). Pruning shears were used to
cut through the pectoral spine and a knife employed to separate the most anterior
spine from adjacent secondary spines and the supporting membrane of the fin.
Collins and Smith (1996) demonstrated that full removal of the primary ray (spine) of
the pectoral fin had no deleterious effect on either the growth or survival of shortnose
sturgeon. Spines were stored and numbered to allow cross-reference with collection
data. Following removal of the spine section, the remaining spine was disinfected
with iodine prior to releasing the fish. Effort was made to minimize handling times
(i.e., time outside of staging or recovery pens); fish were allowed to recover from the
capture process before examination and spine removal. All individuals were released
alive and in apparently good condition.
Laboratory Methods
Spine samples were dried under a fume hood for a period of at least three
weeks. Excess flesh was either mechanically removed with a knife or allowed to
decay via microbial activity. Spine samples were then glued to an epoxy foundation
and sectioned along the transverse plane to 1-2 mm thickness using a Buehler Isomet
17
® low-speed saw. Sections were then fastened to petrographic glass microscopy
slides with thermoplastic glue. If required to improve visual contrast of annular
zones, sections were ground using wetted 800-1200 grit sand paper and polished by
hand using a 0.3 µm alumina slurry on mounted polishing felt.
Annuli were identified and enumerated under 105-1350 X magnification using
a stereo-microscope. For the purpose of our study, we defined an annulus as a
bipartite structure of alternating opaque and translucent bands when viewed under
reflected light (Stevenson & Secor 1999). Narrow translucent zones were assumed to
form during the winter/spring month, whereas wide opaque zones were assumed to
represent growth during the summer/fall feeding periods (Brennan & Cailliet, 1991,
Rien & Beamesderfer 1994). To distinguish “true” from supernumary annuli, only
those annuli were tallied in which the translucent zone formed a distinct continuous
band throughout the posterior lobes of the section (Figure 5). Live-feed video was
used for training purposes, providing a real-time image of spine sections that allowed
for simultaneous interpretations of annuli by multiple readers. A digital camera in
conjunction with imaging software was used to aid in ageing, cataloging and
referencing spine samples.
Age Precision Tests and Age Validation
All samples were aged twice without knowledge of sturgeon size, date of
capture, sampling location, or prior age estimates. Between each age estimation
round, samples were randomized to decrease the likelihood of individual spines being
recognized. Age validation techniques included tests of precision and accuracy.
Within-reader tests of precision were conducted on a test-sample of 55 spines,
18
Opaque zone
Translucent zone
True annulus
False annulusSecondary ray
Opaque zone
Translucent zone
True annulus
False annulusSecondary ray
Figure 5. Posterior lobe from a mounted shortnose sturgeon pectoral spine (transverse section) photographed under reflected light conditions at 450 X magnification. The opaque and translucent zones from a single annulus are indicated. Within the dashed box, a false or ‘supernumary’ annulus is differentiated from the ‘true’ annulus.
19
randomly chosen from a total of 579 specimens. This same test-sample was used to
assess whether precision and bias remained constant during the course of assigning
ages for field-collected sturgeons. Precision was evaluated with a paired t-test,
testing the hypothesis that differences between paired interpretations of the same fin
spine = 0. Age estimates derived during the primary ageing exercise (n = 579) were
compared with those from the precision test for temporal drift in my ageing
technique. An age bias plot and an age frequency table were constructed to visualize
within-reader precision (Campana et al. 1995).
Precision was assessed using two indices of bias, Average Percent Error
(APE) (Beamish & Fournier 1981) and the Coefficient of Variation (CV) (Chang
1982). The APE was calculated as:
∑=
−⎟⎠⎞
⎜⎝⎛×=
R
i XjXjXij
RAPE
1
1100
and CV was calculated as:
( )
Xj
RXjXij
CV
R
i∑= −
−
×= 1
2
1100
where R is the number of reads per sample, Xij is the ith age determination of the jth
fish, and Xj is the mean estimated age of the jth fish. Both methods yield a single
index value for each spine, which then was averaged across all spines to generate a
mean index of precision (Campana et al. 1995). Index values were used to evaluate
whether (1) ‘drift’ or gradually shifting age determinations occurred over the period
of analysis, (2) studies from the literature reported similar error rates in age
interpretations, and (3) systematic biases, such as increased error with larger and
20
older fish, occurred (Campana et al. 1995). Drift was evaluated by grouping
individual CV values into four (each numbering 138 or 139) bins, which were arrayed
in time across the period of age determinations. A Kruskal-Wallace non-parametric
test of means was used to test for a temporal shift in CV across the bins. Age-based
systematic bias was also investigated using a Kruskal-Wallace non-parametric test of
means. Spines were binned into four successive categorical age classes for the latter
analysis.
Ageing accuracy is a measure of “correctness”, the error present between an
estimate and the actual value. Accuracy is a problem common to ageing studies of
wild fish; samples of hard structures (spine, otolith, scale) from known-age
individuals are rare. First, I tested the hypothesis that annuli are deposited at a yearly
rate by conducting a marginal increment analysis (MIA) (Haas & Resnick 1995,
Campana 2001). This technique evaluates the seasonal progression of annulus
formation by measuring the opaque zone deposited after the last identifiable
translucent zone at the margin of the fin spine section (Stevenson & Secor 1999).
MIA utilizes the marginal increment ratio (MIR):
AMIMIR 1
×=
where MI is the width of the outermost opaque zone, measured from the most distal
translucent zone to the edge of the section, and A is the mean width of the three annuli
deposited prior to the marginal increment (Figure 6).
Width of marginal increments were measured using digital analysis software
(ImageJ ©). For the mean monthly MIRs, I chose 25 samples at random for
21
Third annulus
Second annulus
Penultimate annulus
Marginal increment
Transect
Core region
Third annulus
Second annulus
Penultimate annulus
Marginal increment
Transect
Core region
Figure 6. Posterior lobe of a shortnose sturgeon pectoral spine (36 X magnification) with annuli enumerated according to their position relative to the outermost annulus (marginal increment). Annuli are demarcated by the circles and increment width for each annulus is indicated by the brackets.
22
increment analysis with the exception of the August sample, in which the 25 spines
represented the entire sample for that month. Mean MIR values for each sampling
period (c. 2 month intervals) were calculated and analyzed for trends in annulus
formation. The monthly progression of mean MIR values were evaluated with a
series of pair-wise means comparisons (Tukey-adjusted). Following procedures to
diagnose the assumptions of ANOVA, mean monthly MIR values were grouped by
season (Spring, Summer and Fall) to correct for heterogeneity of residual variance
between months. March and April samples were classified as Spring, June and
August samples as Summer, and both November samples (2003 & 2004) were
combined into a Fall sample.
To evaluate the accuracy of individual ages, 59 pectoral spines were collected
from known-age, hatchery-raised shortnose sturgeon from three facilities. Twenty-
six spines (4-9 years old) were obtained from the USFWS Warm Springs National
Fish Hatchery in Warm Springs, GA; 25 spines (3-20 years old) were collected from
the USFWS Bears Bluff National Fish Hatchery on Wadmalaw Island, SC; and 8
spines (8-20 years old) were collected from the Fisheries and Aquatic Sciences
facility of the University of Florida, Gainesville, FL. As age data were not available
for 13 of the 59 spines, only 46 spines were used in this analysis. Spines were
removed from the hatchery fish, then sectioned and annuli enumerated as described
above for field-caught specimens. Comparisons of known and estimated ages were
made using a paired t-test and linear regression analysis. The effects of known age
and hatchery source on accuracy were investigated using a Kruskal-Wallace non-
parametric test of means. To evaluate known-age effect, samples were classified into
23
young (1-6 years) or old (7-20 years) age groups to satisfy the requirement of
homogeneity of variance between groups.
Growth
I derived age estimates from pectoral spines to construct mixed-sex growth
models using the von Bertalanffy (1938) and Gompertz (Weatherley & Gill 1987)
growth equations. Least-squares regression was used to fit linear, exponential and
curvilinear functions to the size-at-age data to verify the most parsimonious model.
Only sturgeon of estimated age 5 - 23 years were included in the growth models. By
limiting the ages of sturgeon included in the analysis I was able to model growth
based upon age-classes for which multiple observations were available. Growth
models were fit to mean size-at-age for TL (mm) and weight (kg). Using mean size-
at-age values prevented more abundant age-classes in the catch data from
disproportionately influencing parameter estimates. I assessed fit based on the
relationship of predicted growth curves to actual data through visual examination of
residuals and the coefficient of determination.
von Bertalanffy Growth Model
Parameters for the growth model were derived through an iterative procedure
(Excel, Solver ©) that minimized the least squares of predicted minus the observed
length-at-age. The TL-based von Bertalanffy growth model is described by the
equation:
[ ])0(exp1 ttKt TLTL −−
∞ −=
24
where TLt is the mean predicted length (mm) at age t (years), TL∞ is the average
asymptotic length of the species, t0 is the x-intercept corresponding to the predicted
age at which size is 0 mm, and K is the Brody growth coefficient (Ricker 1975).
Predicted weight-at-age from the von Bertalanffy growth model required a
preliminary calculation of the allometric growth parameter b (Hillborn & Walters
1992), according to the following equation:
( )bt TLaw =
which was fitted by regression. The von Bertalanffy growth equation for weight at
age is:
⎥⎦
⎤⎢⎣
⎡−= −−
∞bttK
t WW )0(exp1
where Wt is the mean predicted weight (kg) at age t (years), W∞ is the average
asymptotic weight (kg) of the species, and t0 and k are defined as per the length-at-
age model above (Ricker 1975).
Initial estimates of growth parameters were calculated using Ford-Walford
plots (Ricker 1975). Subsequent iterative solving of parameters used mean size-at-
age to predict model parameters.
Gompertz Growth Model
The Gompertz growth model is typically successful at representing growth in
weight and represents a more sigmoidal shape than the von Bertalannfy growth
model, though it can also be applied to length-at-age data (Ricker 1975). The
Gompertz model used to describe growth in weight:
25
)exp1(0 exp
gtGt ww
−−=
where wt is weight at age t, w0 is the predicted weight at age t0, G is the instantaneous
growth rate at age t0, and g describes the rate at which G decreases over time. The
model was the same for length-at-age:
)exp1(0 exp
gtGlTL−−=
with operational definitions of the parameters identical to those for weight.
General Growth Models
Three other models were fit to the size-at-age data for both weight and TL to
describe shortnose sturgeon growth stanzas. The generalized growth models were
linear, exponential, and curvilinear (quadratic):
αββ
α
αββ
++=
=
+=
xxy
y
xyx
12
2
exp
)(
As in the case of the von Bertalanffy and Gompertz models, these generalized growth
models were fit to the mean size-at-age data through least-squares regression. The
generalized growth models were applied across the same interval of age-classes (5-23
years) as the von Bertalanffy and Gompertz models.
Mark-recapture
Original capture data were obtained from the USFWS (Maryland Fishery
Resources Office, Annapolis, MD) for sturgeon with identification tags (e.g., PIT,
Floy, Carlin) from previous studies. It was sometimes necessary to predict TL (mm)
and weight (kg) data points from the available Fork Length (FL) data (Table 3).
26
Table 3. Tag and meristic data of previously marked shortnose sturgeon recaptured during field sampling (from USFWS sturgeon tagging database). FL0 = Fork Length at tagging, FLR = Fork Length at recapture, TL0 = Total Length at tagging, TLR = Total Length at recapture, Wt0 = weight at tagging, WtR = weight at recapture.
2.48
2.5
*2.9
2.32
2.76
3.5
*2.47
3.55
2.02
1.36
*1.8
2.07
3.4
2.18
*2.16
2.12
*1.64
1.6
2.36
2.58
Wt0(kg)
630
940
832
795
865
947
856
985
780
799
746
745
900
820
735
811
655
715
839
800
TLR(mm)
630
790
715
695
733
809
727
852
670
686
640
650
758
710
629
705
555
635
721
683
FLR(mm)
1.356255454139357F3F
4.4815675410A540054
3.05*8096952270450803
2.4*767660#6712 & #671122022F7652
3.78156854138575F49
4.7*857735#5266 & #53641F4875574D
4*767660227D170F56
4.58557302223761752
2.4*713615227D614015
2.3*75765241455C147B
2.2*701605223342765A
3705620413931453C
3.3*821705#5748 & #57471F485E0338
2.7752645410A5D414D
2.2*731630#4814 & #4813
2.8*773665#7162 & #716122021C7240
1.8*656568#31611F3D6B0732
2.1*659570221E2C707F
3.2*78467441132D5C1E
37606452224160E70
WtR(kg)
TL0(mm)
FL0(mm)
FloyTag #(s)PIT Tag #
6/3/1996
11/14/1996
11/21/1995
3/15/1995
5/22/1996
12/6/1994
11/21/1995
3/20/1996
11/27/1995
5/9/1997
4/22/1995
4/9/1996
12/7/1994
11/14/1996
4/15/1994
3/17/1995
4/12/1994
5/5/1995
7/2/1997
3/18/1996
Tagging date
*TL or weight extrapolated from FL, transformations based on relationships calculated from the current study.
2.48
2.5
*2.9
2.32
2.76
3.5
*2.47
3.55
2.02
1.36
*1.8
2.07
3.4
2.18
*2.16
2.12
*1.64
1.6
2.36
2.58
Wt0(kg)
630
940
832
795
865
947
856
985
780
799
746
745
900
820
735
811
655
715
839
800
TLR(mm)
630
790
715
695
733
809
727
852
670
686
640
650
758
710
629
705
555
635
721
683
FLR(mm)
1.356255454139357F3F
4.4815675410A540054
3.05*8096952270450803
2.4*767660#6712 & #671122022F7652
3.78156854138575F49
4.7*857735#5266 & #53641F4875574D
4*767660227D170F56
4.58557302223761752
2.4*713615227D614015
2.3*75765241455C147B
2.2*701605223342765A
3705620413931453C
3.3*821705#5748 & #57471F485E0338
2.7752645410A5D414D
2.2*731630#4814 & #4813
2.8*773665#7162 & #716122021C7240
1.8*656568#31611F3D6B0732
2.1*659570221E2C707F
3.2*78467441132D5C1E
37606452224160E70
WtR(kg)
TL0(mm)
FL0(mm)
FloyTag #(s)PIT Tag #
6/3/1996
11/14/1996
11/21/1995
3/15/1995
5/22/1996
12/6/1994
11/21/1995
3/20/1996
11/27/1995
5/9/1997
4/22/1995
4/9/1996
12/7/1994
11/14/1996
4/15/1994
3/17/1995
4/12/1994
5/5/1995
7/2/1997
3/18/1996
Tagging date
6/3/1996
11/14/1996
11/21/1995
3/15/1995
5/22/1996
12/6/1994
11/21/1995
3/20/1996
11/27/1995
5/9/1997
4/22/1995
4/9/1996
12/7/1994
11/14/1996
4/15/1994
3/17/1995
4/12/1994
5/5/1995
7/2/1997
3/18/1996
Tagging date
*TL or weight extrapolated from FL, transformations based on relationships calculated from the current study.
27
Transformation of supplied FL to TL (n = 13) was carried out using a TL-FL
relationship generated from our dataset:
FLTL *16.1=
(n = 580; R2 = 0.94) for this purpose. It was also necessary to estimate sturgeon
weight at the time of original tagging from FL data (n = 6). Weight was predicted
from FL by the equation:
( )FLWt *0045.0exp127.0=
(n = 580; R2 = 0.80). Growth increment over the interval between tagging and
recapture was adjusted to yearly growth increments for both TL and weight
responses. Back-calculated ages at tagging were derived from age determinations
(hard part analysis) and independently predicted using the von Bertalanffy growth
model for TL applied to the size at initial tagging. The two sets of age-at-tagging
values were tested for significant differences (Paired t-test). Magnitude of
incremental growth was tested for size-based trends in growth trajectories. Sturgeon
were sorted by TL at recapture, larger fish were defined ≥ 800 mm TL (n = 11) and
smaller fish < 800 mm TL (n = 10). Paired and unpaired t-tests were conducted to
investigate the relationship between sturgeon size and back-calculated mean annual
growth rate.
Gear Selectivity
Gill nets are highly size-selective and introduce bias into demographic
estimates that assume random sampling of a population (Hamley 1975). Therefore, it
was necessary to adjust catch data based on the individual selectivity of each mesh
28
size (10.2, 15.2 and 17.8 cm) used during field captures. Mesh selectivity values
were estimated through manipulation of the selection equation:
mmlmlmlml EsPNqC ,,, =
where C l,m is the catch of a given length class l by mesh size m, ql,m is the proportion
of the population of length class l that is vulnerable to the mesh (catchability), Nl is
the population of a length class l available to the mesh, Pm is the fishing power or
efficiency of the mesh at retaining fish of size l, sl,m is the selectivity of m and
assumed to be size-dependent only, and Em is the effort under which m is deployed
(Hovgård & Lassen 2000).
Direct selectivity estimates are generated through mark-recapture studies or
sampling populations of a known size distribution (Hamley 1975, Gulland 1983).
While conducting field sampling, I attempted alternating fin clips of sampled
sturgeon (right versus left) based on mesh size in which fish were captured to yield a
direct estimate of the selectivity for each mesh size. However, only a single
individual of the 587 total sturgeon captured was identified as a recapture. This was
likely due to the high abundance of shortnose sturgeon present. I therefore employed
indirect methods for estimating mesh selectivities.
Estimates of indirect gillnet selectivity were generated using the Baranov,
Holt and Regression models (Hamley 1975, Hovgård & Lassen 2000). Models were
evaluated based on the agreement between predicted size-selectivity curves of catch
versus the observed catch data. To minimize bias in estimates of gear selectivity,
only catch data from sampling periods in which all nets were fished at the same
location and day were used. Catch included in the analysis was therefore limited to
29
40-45 net minute deployments at the Esopus Meadows (March) and Albany (April)
sampling locations. By limiting the data to these sampling periods, Em was
standardized among meshes as was the population Nl available to the gear. This also
ensured physical river conditions (e.g., turbidity) were similarly affecting Cl,m.
Predicted selection curves were visually examined for a representative fit to
the Esopus Meadows data, then plotted against the entire catch and again inspected
for fit. The coefficient of determination (Ott & Longnecker 2001) between observed
and predicted values was calculated for each model as an indicator of fit.
Baranov Model
The central assumption of the Baranov model (Baranov 1914) is the ‘principle
of geometric similarity’ and states that selectivity between meshes sizes is
proportional due to the geometric similarity of mesh construction and the
morphological similarity of different sized fish of the same species (Hamley 1975).
This implies that selectivity, s, is constant across all combinations of meshes and fish
size for which the ratio of fish length to mesh size is the same:
( ) ),(, mklksmls ∗∗=
where k is a constant (Hamley 1975). Given this assumption, it follows that each
net (i.e., mesh size) is equally efficient at capturing some optimal length of fish, lo,
dictated by the size of the mesh (Hovgård & Lassen 2000). The catch equation can
then be rearranged to the form:
⎥⎦⎤
⎢⎣⎡∗∗∗=mlsNlq
EC
lm
ml,
30
where the ratio of catch per unit effort (Cl,m Em-1) is proportional to the selectivity, s(l
m-1).
The NORMSEP procedure of FiSAT II © (FAO-ICLARM Fish Stock
Assessment Tools 2000) was used to generate selectivity curves based on the
assumptions of the Baranov model (Hovgård & Lassen 2000). This procedure used
maximum likelihood estimates to separate length-frequency bins for each mesh into
normally distributed selectivity curves. Selection curves were scaled to a common
maximum value of 1 (Hamley 1975).
Holt Model
The Holt model (Holt 1963) does not rely on the principle of geometric
similarity; rather it applies a standard linear regression to the catch data and assumes
selectivity curves for each mesh follow a Gaussian distribution and have the same
variance σ2 (Hovgård & Lassen 2000). The selection model is:
( )⎥⎥⎦
⎤
⎢⎢⎣
⎡ −−= 2
2,
,2
expσ
kmCs ml
ml
where k is a selection factor equal to 1/K with K = lo / m, and σ2 is the variance
associated with the curve (Hamley 1975, Hovgård & Lassen 2000). Holt proposes
that the logarithmically transformed catch ratio of a given length class between mesh
sizes is proportional to the similarly transformed selectivity values for that length
class:
⎟⎟⎠
⎞⎜⎜⎝
⎛=⎟
⎟⎠
⎞⎜⎜⎝
⎛
2,
1,
2,
1, lnlnml
ml
ml
mlss
CC
31
This method used a least squares linear regression of the ln(Cl,m1 Cl,m2-1) values
plotted against the mode of each length class. The following equations provided the
y-intercept (β) and slope (α):
2
21
22
2
2)(
σβ
mmk −=
( )2
21
σα
mmk −=
Substituting and solving for the selection factor and variance (parameters k and
σ2) yielded:
( )21
2mm
k−
=α
β
[ ][ ]21
2212 2
mmmm
+−
=α
βσ
from which parameter values selectivity curves were generated.
Regression Model
The regression method used a matrix-based framework that employed a power
transformation of the catch data followed by least-squares regression to account for
the error structure (Hovgård 1996). A derivation of the selection equation yields an
estimate of the population per size class:
( )
( )
β
β
β1
2,
,,
1
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
=∑
∑
mmlmm
mmlmmml
sEP
sPECqN
32
This equation provided the least-squares estimate for qNl values, which were then
minimized through an iterative procedure that adjusted the qNl’s to yield the least-
squares sum (Hovgård & Lassen 2000):
( )[ ]∑∑ −=m l
lmlmmlm qNsPECLsq2
,,ββ
The flexibility of the regression framework lies in the mutability of the model,
providing a uniform method of describing a range of error structures (e.g., high to low
contagion) (Hovgård & Lassen 2000). Selection equation parameters Pm and Em can
be ignored if catch is limited to periods of equal effort per mesh and fishing power is
the same among gears (Pm1 = Pm2 = … = Pmn = 1).
A log-normal selection curve was assumed, with error terms following a
Poisson distribution (β = 0.5):
⎟⎟⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜⎜⎜
⎝
⎛
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛⎟⎟⎠
⎞⎜⎜⎝
⎛−⎟
⎠⎞
⎜⎝⎛−
=2
2
,
lnln*5.0
expσ
kml
s ml
Non-linear least squares regression of the residuals (predicted – observed catch) was
then used to solve for the selection factor k and the variance structure σ2.
Recruitment
Following correction of actual catch for gear selectivity, catch values for all
age-classes across meshes were subjected to an effort modifier:
⎟⎟⎠
⎞⎜⎜⎝
⎛=
m
total
seladj
EEC
C
33
where Cadj is the catch adjusted for gear selectivity and effort, Csel is the partially
adjusted catch (gear selectivity only), Etotal is the total fishing effort (net minutes) and
Em is the mesh specific fishing effort (net minutes).
Adjusted catch estimates, Cadj, for each estimated age-class were summed
across meshes and the resultant age structure was subjected to a catch curve analysis
(Ricker 1975). Adjusted catch values were logarithmically transformed (y-axis) and
plotted against their corresponding age-classes along the abscissa. An estimate of the
annual instantaneous mortality rate, Z, was generated by conducting a least-squares
linear regression through the descending portion of the plot where slope = Z.
Analysis was limited to ages that had fully recruited to the gear. The presence of
anomalous year-class strength, indicated by the deviation of adjusted catch-at-age
from the predicted catch, was analyzed by applying ANOVA to the least-squares
catch-curve residuals. Residuals were grouped into four successive stanzas composed
of five year-classes each (e.g., Group1 = 1999-1995, etc).
The exponential decay model was then used to predict the relative strength of
each year-class at age = 1 year (yearlings) from the adjusted catch data:
Ztt NN −= exp1
where Nt is the cohort population at time t, N1 is the cohort population at t = 1 year of
age, and Z is the estimate of the total annual mortality rate. Estimated age at capture
was standardized to reflect a sampling date of November 2004 for all catch data (i.e.,
estimated ages for sturgeon captured during November 2003 were augmented by one
year). Low catch numbers occurred for the oldest age-classes, so analysis was
limited to age 5-26 years (1999-1979 year-classes). A recruitment strength index
34
(RSI) was calculated by assuming a constant Z = -0.22 and standardizing the resultant
year-class strengths to a relative scale of 1. Year-class abundances at the transition to
age 1 (yearlings) were estimated for the years 1979-1999, by apportioning 50,000
(approximate Hudson River population estimate, Bain et al. 2000) across the
observed age class frequencies (corrected for gear selectivity and cumulative natural
mortality). To account for the population fraction attributable to age-classes that
were not sampled (i.e., 1-4, 24, 27 & 28 years), predicted abundances were generated
assuming a constant annual mortality of Z = 0.22. Year-class assignments to
abundance estimates were made by subtracting age from year of capture and were
assumed to represent annual cohort strength of yearlings.
Fall Juvenile Survey
Bycatch data (1985-2003) from the ongoing Hudson River Utilities Fall
Juvenile Survey were obtained with permission from Dr. Mark Mattson, Normandeau
Associates. Gear used in the survey is a benthic 3 m beam trawl of 3.8 cm body mesh
with a cod end and cod end liner of 3.2 and 1.3 cm respectively (Geoghegan et al.
1992). As a method of corroborating my hindcast year-class strength estimates, I
compared trends in RSI estimates with shortnose sturgeon bycatch CPUE (catch 1000
m3 -1) from the trawl survey. Trawl survey data was iteratively lagged in a succession
of single-year time steps. The lagged time step with the highest regression correlation
coefficient was considered the most representative fit.
35
Environment: Flow
Historical annual flow data (years: 1973-1997, 2000) from the US Geological
Survey’s Green Island, NY monitoring station (# 01358000) located at the Federal
Dam in Troy was downloaded from the US Geological Survey website (USGS 2005).
The dataset contained one mean flow volume datum (ft3 sec-1) per month of each
calendar year downloaded, except for 1997 which only included flow volumes for
January though September. Ambient water temperature was also included in the
dataset; unfortunately the gaps in temperature data were quite pervasive and
precluded the inclusion of water temperature as a variable in this analysis. The
correlation of hydrographic conditions on RSI values of age 1 shortnose was
evaluated using non-parametric correlation analysis (Spearman’s rank-correlation
coefficient) due to persistent non-normality of the data (Zar 2004). Analyses were
constructed to identify correlations between flow volumes and both adult pre-spawn
conditioning and early life history stages that may influence the success of annual
recruitment.
Specifically, I tested for correlations between flow and the conditioning of
pre-spawn adults as defined by the predicted strength of the following year’s age 1
year-class. This was conducted by comparing monthly flow volumes from October to
December to the following year’s relative recruitment strength. This analysis was
extended through January to April of the same year for which the recruitment
estimates were developed. Secondly, variability in flow volume was investigated in
terms of larval survival and subsequent influence on final yearling recruitment values.
Flow from the months of May and June were compared to same-year recruitment
36
index values. Third, I tested for correlations between a given year’s final recruitment
strength and the concurrent summer, fall and winter (July-December) flow conditions.
This provided a means of looking at flow as it related to the YOY juvenile period and
the final recruitment strength index. Monthly flow volumes were compared singly,
grouped by 2-3 month intervals and grouped by the entire range of values (pre-spawn
adult only).
37
Results
Field Sampling
A total of 587 shortnose sturgeon were captured from three different sampling
locations on the Hudson River. We captured 341 fish at Esopus Meadows during
November 2003 and March 2004, targeting over-wintering and pre-spawn
congregations. A total of 129 shortnose sturgeon was sampled during the spring
spawning event near Albany in April 2004. Summer sampling yielded 83 individuals
in the Catskill area during the months of June and August 2004. The final 43 fish
were taken in November 2004, again from the over-wintering location near Esopus
Meadows. Catch per unit effort (CPUE) was greatest during spring and fall sampling
periods with a mean CPUE ± s.e. across months of 1.9 ± 0.9 fish minute-1 (Figure 7
A). Different meshes showed variable CPUE, with the 15.2 cm mesh yielding the
highest mean CPUE across sampling months (0.33 fish minute-1). The CPUE for the
10.2 cm and 17.8 cm meshes were lower at 0.13 fish minute-1 and 0.2 fish minute-1,
respectively (Figure 7 B).
Size data for the total catch followed Gaussian distributions. The distribution
of weights was slightly skewed with a mean of 2.78 ± 0.04 kg (s.e.). The length
distribution was left skewed with mean TL = 783 ± 4 mm (Figure 8). The modal age
of the sample was 13 years with a maximum estimated age of 30 years (Figure 9).
The largest individual captured was 1045 mm TL and weighed 9.0 kg. The smallest
sturgeon measured 490 mm TL (0.5 kg) and was one of several estimated at 5 years
of age.
38
0
50
100
150
200
250
Nov Mar Apr Jun Aug Nov
Cat
ch (n
umbe
r)
0
2
4
6
8
10
12
14
CPU
E (fish net-minute -1)
0
2
4
6
8
Nov Mar Apr Jun Aug Nov
Sampling month
CPU
E (fi
sh n
et-m
inut
e-1)
10.2 cm mesh15.2 cm mesh17.8 cm mesh
CPUE
B
A
0
50
100
150
200
250
Nov Mar Apr Jun Aug Nov
Cat
ch (n
umbe
r)
0
2
4
6
8
10
12
14
CPU
E (fish net-minute -1)
0
2
4
6
8
Nov Mar Apr Jun Aug Nov
Sampling month
CPU
E (fi
sh n
et-m
inut
e-1)
10.2 cm mesh15.2 cm mesh17.8 cm mesh
CPUE
B
A
Figure 7. Catch per unit effort (CPUE) of Hudson River shortnose sturgeon over the course of the sampling period (2003-04). CPUE for each sampling period by gear type is shown in panel.
X = 2.8 ± 1.02 (s.d.) kg
X = 783 ± 86 (s.d.) mm
Weight (kg)
0
1020304050607080
440
480
520
560
600
640
680
720
760
800
840
880
920
960
1000
1040
Total Length (mm)
Cat
ch (n
umbe
r)
0
25
50
75
100
125
150
0 1 2 3 4 5 6 7 8 9
Cat
ch (n
umbe
r)A
B
X = 2.8 ± 1.02 (s.d.) kg
X = 783 ± 86 (s.d.) mm
Weight (kg)
0
1020304050607080
440
480
520
560
600
640
680
720
760
800
840
880
920
960
1000
1040
Total Length (mm)
Cat
ch (n
umbe
r)
0
25
50
75
100
125
150
0 1 2 3 4 5 6 7 8 9
Cat
ch (n
umbe
r)A
B
Figure 8. Total catch (n = 587) of shortnose sturgeon over the course of the sampling period (2003-04) by A) weight (kg) and B) Total Length (mm).
Estimated age (years)
Num
ber o
f sho
rtnos
e st
urge
on
40
10
2030
40
50
60
7080
90
100
2 6 8 10 12 14 16 18 20 22 24 26 3028
X = 13 years ± 3 (s.d.)
Figure 9. Distribution of estimated ages from interpretable spines (n = 554).
41
Age Precision Tests and Age Validation
After eliminating damaged or unreadable spines (n = 25), age estimates were
produced for 554 pectoral spines. Spines were determined to be unreadable due to a
suite of optical features that impeded age estimation (Figure 10). Supernumary
annuli were present in most sections, caused by division of a single annulus into two
or multiple lamellar structures. Spans of very narrowly spaced annuli, also termed
‘banding’ were observed; investigators have speculated that banding is associated
with energy deferred from somatic growth to gonadal development in the years prior
to a spawning event (Cuerrier 1951, Roussow 1957). Annular width was variable in
most sections with a typical pattern of narrow multiple annuli near spine edges. The
inclusion of secondary rays embedded within the primary spine was common though
inclusions were typically easy to identify. Resorption or deposition of calcareous
material and physical deterioration of spines were two primary reasons for rejecting
sections and can pose a significant impediment to accurate interpretation of annuli.
Repeated age estimates were relatively precise with 40% of the spines
assigned the same age in each round, 43% of the estimates ± 1 year, 14% ± 2 years,
and 3% of the estimate ± 3 years or more. The greatest discrepancy encountered was
8 years, which occurred in a single specimen. Precision was high with an APE of
3.0% and a CV of 4.0% for the entire sample, and an APE of 6.0% and CV of 8.0%
for the test subsample of 55 spines. Non-parametric tests of means for CV values
binned by age-group indicated the absence of age related bias in precision (Kruskal-
Wallace χ2 = 5.8, p = 0.12). Similarly, there was no significant difference in mean
42
BA
C D
BA
C D
Figure 10. Transverse sections of pectoral spines viewed under reflected light microscopy. Common sources of ageing error are indicated by arrows in each image. A) Spine showing initial stages of resorbtion of calcified material, canalization and proteinaceous deposits. B) Inclusion of secondary rays within the posterior lobes; inclusion can vary from two, one or none within the primary spine. C) Inconsistent annulus widths. D) “Crowding” of annuli near edge of spine.
43
CV among groups binned by the order of interpretation (i.e., temporal bias or
drift) (Kruskal-Wallace χ2 = 2.8, p = 0.43).
Analysis of the test subsample (n = 55) indicated that interpretation of annuli
was variable though no temporal bias was apparent over the course of the ageing
study. A paired t-test indicated that age estimates from the test sample were not
significantly different from initial to final estimates (t = 1.67, p = 0.28). A visual
assessment of the age-bias plot indicated that my age interpretations were unbiased
for shortnose sturgeon with estimated ages 5 to 30 years (Figure 11). Age estimation
was very precise for sturgeon estimated at 5 to 16 years of age, with the 95%
confidence intervals associated with these age-classes estimates overlapping the 1:1
ratio line. Precision was substantially less for sturgeon aged 17 and older with the
95% confidence intervals for five of the nine highest age-classes falling outside the
1:1 ratio line.
Age estimates for hatchery-reared shortnose sturgeon did not accurately
correspond to known ages (Table 4). Although a nonparametric means comparison
failed to find a significant difference between the estimated and actual ages (Z = -
1.56, p = 0.12), a visual examination of the residuals indicated a systematic bias
(Figure 12 A). Regression analysis of estimated ages on actual ages indicated a
significant deviation from the 1:1 relationship, signifying a bias in age interpretations
(F = 32.11, p < 0.0001) (Figure 12, B). Similarly, contingency table analysis showed
a significant age effect on accuracy (χ2 = 9.94, p = 0.002), with the difference
between estimated and actual ages varying significantly between young (1-6 years)
and older (7-20 years) fish. Estimated ages of younger fish typically fell above the
44
0
5
10
15
20
25
30
35
0 5 10 15 20 25 30 35
First age estimates (years)
Seco
nd a
ge e
stim
ates
(yea
rs)
0
5
10
15
20
25
30
35
0 5 10 15 20 25 30 35
First age estimates (years)
Seco
nd a
ge e
stim
ates
(yea
rs)
Figure 11. Age-bias plot of subsampled spines (n = 55) for Hudson River shortnose sturgeon. Mean second ages are plotted with 95% confidence intervals. The 1:1 ratio line is shown.
Table 4. Statistical analyses used to diagnose accuracy of age determinations for hatchery raised sturgeon.
Reject null hypothesis: Significant difference between young (3-8 years) and older fish (14-20 years)
0.0001Z = -3.8H0 : accuracy of age estimates are unaffected by age of sturgeon
WilcoxonNonparametric rank sum test
χ2 = 3.03
Z = 1.56
F = 31.29
t = 0.02
Test statistic
0.22
0.12
0.0001
0.99
p value (α > 0.5)
Non-parametric
Parametric
Reject alternative hypothesis: No significant difference in accuracy among hatcheries
H0 : accuracy of age estimates are unaffected by hatchery of origin
Kruskal-Wallis
Reject alternative hypothesis: No significant difference between estimated and true ages
H0: no difference between estimated and true ages
WilcoxonNonparametric rank sum test
Reject null hypothesis: Slope of residuals is significantly different than zero
H0 : accuracy of age estimates are unaffected by age of sturgeon
Linear regression
Reject alternative hypothesis: No significant difference between estimated and true ages
H0: no difference between estimated and true ages
Paired t-test
ConclusionsNull hypothesisStatistical test
Reject null hypothesis: Significant difference between young (3-8 years) and older fish (14-20 years)
0.0001Z = -3.8H0 : accuracy of age estimates are unaffected by age of sturgeon
WilcoxonNonparametric rank sum test
χ2 = 3.03
Z = 1.56
F = 31.29
t = 0.02
Test statistic
0.22
0.12
0.0001
0.99
p value (α > 0.5)
Non-parametric
Parametric
Reject alternative hypothesis: No significant difference in accuracy among hatcheries
H0 : accuracy of age estimates are unaffected by hatchery of origin
Kruskal-Wallis
Reject alternative hypothesis: No significant difference between estimated and true ages
H0: no difference between estimated and true ages
WilcoxonNonparametric rank sum test
Reject null hypothesis: Slope of residuals is significantly different than zero
H0 : accuracy of age estimates are unaffected by age of sturgeon
Linear regression
Reject alternative hypothesis: No significant difference between estimated and true ages
H0: no difference between estimated and true ages
Paired t-test
ConclusionsNull hypothesisStatistical test
46
5
10
15
20
0 5 10 15 20
Actual age (years)
Estim
ated
age
(yea
rs)
-15
-10
-5
0
5
10
15Pa
ired
diff
eren
ces
(est
imat
ed -
true
ages
, yea
rs) A
B
5
10
15
20
0 5 10 15 20
Actual age (years)
Estim
ated
age
(yea
rs)
-15
-10
-5
0
5
10
15Pa
ired
diff
eren
ces
(est
imat
ed -
true
ages
, yea
rs) A
B
Figure 12. Analysis of age determinations of hatchery-raised shortnose sturgeon (n = 46). A) Paired differences of estimated minus true ages (residuals) plotted against actual age (abscissa) with positive (young fish, 1-6 years) and negative (older fish, 7-20 years) bias clearly visible. B) Estimated ages (y-axis) plotted against actual ages (abscissa) with a linear regression (least squares) line (solid line) relative to the 1:1 line (dotted line).
47
1:1 ratio line indicating a positive bias, while the estimated ages of older fish were less
than the actual values, indicative of a negative bias. The microstructure of hatchery
spines showed wide variation in the optical characteristics of annuli and related structures
in comparison with spines from wild-captured individuals (Figure 13, A-D). Hatchery
spines often showed irregular deposition of calcareous material and blurred or faint
annuli. Physical deterioration and compaction was pronounced in several hatchery
spines. Despite widely varying microstructures among individuals from hatcheries, the
hatchery source itself did not affect the accuracy of age estimates (χ2 = 3.03, p = 0.22).
Mean monthly marginal increment ratios increased steadily from a nadir in March
to a maximum in November (Table 5). When monthly ratio values were grouped by
season, the Spring ratio was significantly less than both the Summer (t = -2.54, p = 0.03)
and Fall ratios (t = 3.8, p < 0.001). Summer and Fall ratios were not significantly
different (t = 1.34, p = 0.38) although the mean ratio value increased from Summer to
Fall (Figure 14, A). Assumptions of the ANOVA model were met under the seasonal
grouping, but not for the monthly comparisons. Due to low numbers within age-classes
during some of the sampling periods and very small sample size for the individuals with
the oldest estimated ages (i.e., ages ≥ 25, n = 1), it was necessary to pool age-classes to
produce composite monthly and seasonal MIR values. The youngest (5-7 years) and
oldest (19-26 years) sturgeon included in the MIA were not present in all months or
seasons. Sturgeon with estimated ages ranging from 12-15 years were included in all
monthly and seasonal means, while each of the three seasonal means contained sturgeon
aged 8 to 18 years. There was no indication
48
B
Age: 6 years old
Facility: USFWS National Fish Hatchery
Location: Warm Springs, GA
Age: 3 years old
Facility: USFWS National Fish Hatchery
Location: Wadmalaw Island, SC
A
C
Age: 17 years old
Facility: USFWS National Fish Hatchery
Location: Warm Springs, GA
Deterioration/ malformation of
spine
Multiple false annuli
D
Spine with irregular annuli formation and
calcareous deposition
Age: 18 years old
Facility: Univ. of Florida, Fisheries and Aquatic Sciences facility
Location: Gainesville, FL
B
Age: 6 years old
Facility: USFWS National Fish Hatchery
Location: Warm Springs, GA
Age: 3 years old
Facility: USFWS National Fish Hatchery
Location: Wadmalaw Island, SC
A
C
Age: 17 years old
Facility: USFWS National Fish Hatchery
Location: Warm Springs, GA
Deterioration/ malformation of
spine
Multiple false annuli
D
Spine with irregular annuli formation and
calcareous deposition
Age: 18 years old
Facility: Univ. of Florida, Fisheries and Aquatic Sciences facility
Location: Gainesville, FL
Figure 13. Transverse sections of pectoral spines from four different hatchery-raised shortnose sturgeon. A & B) depict hatchery spines that appear to be laying down consistent increments without noticeable deformation or artificial structures. C & D) depict spines that are deteriorated, possess multiple false structures and large calcareous deposits.
49
Table 5. Mean Marginal Increment Ratio values (MIR) by sampling month and season ± 95% Confidence interval. Means with any identical letters are not significantly different at �= 0.05 (Tukey-Kramer protected against experimentwise error inflation).
0.85 ± 0.12November (‘03)0.80b ± 0.09Fall
0.75 ± 0.12November (‘04)
0.73b ± 0.09Summer
0.78 ± 0.13August
0.66 ± 0.12June
0.65 ± 0.12April0.58a ± 0.08Spring
0.55 ± 0.07March
Mean seasonal MIR with 95% C.I.
Sampling season
Mean monthly MIR with 95 % C.I.
Sampling month
0.85 ± 0.12November (‘03)0.80b ± 0.09Fall
0.75 ± 0.12November (‘04)
0.73b ± 0.09Summer
0.78 ± 0.13August
0.66 ± 0.12June
0.65 ± 0.12April0.58a ± 0.08Spring
0.55 ± 0.07March
Mean seasonal MIR with 95% C.I.
Sampling season
Mean monthly MIR with 95 % C.I.
Sampling month
0.85 ± 0.12November (‘03)0.80b ± 0.09Fall
0.75 ± 0.12November (‘04)
0.73b ± 0.09Summer
0.78 ± 0.13August
0.66 ± 0.12June
0.65 ± 0.12April0.58a ± 0.08Spring
0.55 ± 0.07March
Mean seasonal MIR with 95% C.I.
Sampling season
Mean monthly MIR with 95 % C.I.
Sampling month
50
0
0.25
0.5
0.75
1
Spring Summer Fall
Season
Mea
n M
IR
a
bb
MIR = 1, Annulus fully formed
Winter
Apparent period of annulus completion and
initiation
0
0.25
0.5
0.75
1
Spring Summer FallWinter
b
a
b
A
B
Mea
n M
IR
0
0.25
0.5
0.75
1
Spring Summer Fall
Season
Mea
n M
IR
a
bb
MIR = 1, Annulus fully formed
Winter
Apparent period of annulus completion and
initiation
0
0.25
0.5
0.75
1
Spring Summer FallWinter
b
a
b
A
B
Mea
n M
IR
Figure 14. Mean seasonal Marginal Increment Ratios (MIRs) with 95 % C.I. A) Overall Marginal Increment Analysis (MIA, ages 5-26). B) Reduced set MIA (ages 10-17). Seasonal MIR values that have different letters above them are significantly different at α = 0.05 (Tukey-Kramer protected against experimentwise error inflation).
51
of bias when the seasonal MIA was repeated using a common age distribution sub-
sample (10-17 years). Mean MIR values indicated a steadily increasing increment width
from Spring (0.58) to Summer (0.73) to Fall (0.82) (Figure 14, B). Summer (Two sample
t-test, t = -1.65, p = 0.05) and Fall (Two sample t-test, t = -3.26, p < 0.001) MIRs
remained significantly greater than the Spring MIR, corroborating the results from the
full sample MIA.
Growth
Growth curves were calculated based on the sizes selectively captured during the
study and therefore reflect growth rates of sub-adults and adults (ages 5-23 years). Sex
determination in the field was unsuccessful for all but three very ripe males, which
expressed milt upon application of pressure to the abdomen. This precluded determining
sex-specific age-structure and growth models. Following application of all growth
models (Figure 15) to size-at-age data, the von Bertalanffy growth function for weight
and length was chosen as the most representative model. The von Bertalanffy growth
curves derived from TL (mm) measurements was
( )( )42.7*07.0exp1*1045 +−−= tTL
with R2 = 0.93, and TL∞ constrained to the maximum observed TL value of 1045 mm.
The von Bertalanffy growth function was also determined using FL (mm) to ease
parameter comparisons among literature values. The resultant growth function for FL
was
( )( )05.7*06.0exp1*925 +−−= tFL
52
y = 194.7*Ln (x) + 278R2 = 0.94
y = -1.4 x2 + 50.3x + 364R2 = 0.95
Tota
l Len
gth
(mm
)
y = -0.01 x 2 + 0.39x - 0.84
R2 = 0.97
0
2
4
6
8
Weight (kg)
A
DC
B
0 10 20 30 40
( ) ⎟⎠
⎞⎜⎝
⎛ −= −− 2.331.4*1.0exp1*0.5 tWt
( )( )42.7*067.0exp1*1045 +−−= tTL
R2 = 0.93
R2 = 0.87
R2 = 0.97
⎟⎠
⎞⎜⎝
⎛=
−− )*16.0exp1(*59.4exp*05.0t
Wt
0 10 20 30Age (years)
R2 = 0.87 500600700800900
10001100
⎟⎠
⎞⎜⎝
⎛=
−− )*43.0exp1(*42.4exp*10t
TL
G H
E F
Age (years)
R
y = 2.1*Ln (x) - 2.572 = 0.97
0
2
4
6
8
0
2
4
6
8
0
2
4
6
8
0
500600700800900
10001100
0
500600700800900
10001100
0
500600700800900
10001100
Von Bertalanffy
Polynomial model
Gompertz
Ln model
y = 194.7*Ln (x) + 278R2 = 0.94
y = -1.4 x2 + 50.3x + 364R2 = 0.95
Tota
l Len
gth
(mm
)
y = -0.01 x 2 + 0.39x - 0.84
R2 = 0.97
0
2
4
6
8
Weight (kg)
A
DC
B
0 10 20 30 40
( ) ⎟⎠
⎞⎜⎝
⎛ −= −− 2.331.4*1.0exp1*0.5 tWt
( )( )42.7*067.0exp1*1045 +−−= tTL
R2 = 0.93
R2 = 0.87
R2 = 0.97
⎟⎠
⎞⎜⎝
⎛=
−− )*16.0exp1(*59.4exp*05.0t
Wt
0 10 20 30Age (years)
R2 = 0.87 500600700800900
10001100
⎟⎠
⎞⎜⎝
⎛=
−− )*43.0exp1(*42.4exp*10t
TL
G H
E F
Age (years)
R
y = 2.1*Ln (x) - 2.572 = 0.97
0
2
4
6
8
0
2
4
6
8
0
2
4
6
0
2
4
6
8
0
500600700800900
10001100
0
500600700800900
10001100
0
500600700800900
10001100
Von Bertalanffy
Polynomial model
Gompertz
Ln model
Figure 15. Growth curves calculated for mean Total Length (mm) and mean weight (kg) on age (years) with range of size values indicated by vertical bars and denoted by circles. Filled circles were used in the analysis, open circles were not. A & B) Loge growth function; C & D) Polynomial (quadratic) growth function; E & F); von Bertalanffy growth function; G & H) Gompertz growth function.
53
with R2 = 0.92, and FL∞ constrained to the maximum observed FL value of 925 mm.
The allometric growth exponent b was calculated as 3.2 (R2 = 0.85) from the catch data.
The von Bertalanffy growth function for weight was
( ) ⎟⎠
⎞⎜⎝
⎛ −= −− 2.331.4*1.0exp1*0.5 tWt
with Wt∞ constrained to ≥ 4.7 kg (95th percentile by weight observed) (R2 = 0.97 ).
The Gompertz function described growth in length (R2 = 0.87) as
)43.0exp1(42.4exp*10t
TL−−=
with l0 constrained at 10 mm to provide a biologically realistic representation of growth.
Constraining w0 to 0.05 kg, the Gompertz growth function for weight (R2 = 0.87) was
)16.0exp1(59.4exp*05.0t
Wt−−=
Though the Gompertz and tertiary growth models showed high correlations (0.86 ≤ R2 ≥
0.97) and appeared to describe size-at-age relationships in an adequate fashion, the von
Bertalanffy model was chosen based on precedent, ease of inter-study comparisons and
the intrinsic biological meaning of the function’s parameters (Figure 16).
During field sampling, twenty-one previously tagged shortnose sturgeon were
captured (one of which was captured a second time). The majority of these fish
possessed passive integrated transponder (PIT) tags (n = 19), though five also carried
external Floy® or Carlin® tags. Age estimates derived from hard part analysis for
recaptures ranged from 9-24 years, yielding original age-at-tagging estimates from 4-
54
0
400
600
800
1000
1200
Tota
l Len
gth
(mm
)
0
2
4
6
8
10
0 5 10 15 20 25 30Age (years)
Wei
ght (
kg)
0 5 10 15 20 25 30 35Age (years)
200
n = 554
TL∞= 1045, k = 0.07, t0 = 7.42 (R2 = 0.93)
Wt∞ = 5.0, k = 0.1, t0 = 4.3 (b = 3.2, R2 = 0.97)
A B
C D
0
400
600
800
1000
1200
Tota
l Len
gth
(mm
)
0
2
4
6
8
10
0 5 10 15 20 25 30Age (years)
Wei
ght (
kg)
0 5 10 15 20 25 30 35Age (years)
200
n = 554
TL∞= 1045, k = 0.07, t0 = 7.42 (R2 = 0.93)
Wt∞ = 5.0, k = 0.1, t0 = 4.3 (b = 3.2, R2 = 0.97)
0
400
600
800
1000
1200
Tota
l Len
gth
(mm
)
0
2
4
6
8
10
0 5 10 15 20 25 30Age (years)
Wei
ght (
kg)
0 5 10 15 20 25 30 35Age (years)
200
n = 554
TL∞= 1045, k = 0.07, t0 = 7.42 (R2 = 0.93)
Wt∞ = 5.0, k = 0.1, t0 = 4.3 (b = 3.2, R2 = 0.97)
A B
C D
Figure 16. Size-at-age data for Hudson River shortnose sturgeon (combined sex data). Length–age data given by open circles (A) with mean length-at-age given by closed circles with fitted von Bertalanffy curve for ages 5-30 (B). Weight–age data given by open circles (C) with mean weight-at-age given by closed circles with fitted von Bertalanffy curve for ages 5-30 (D).
55
16 years (Table 6). Predicted age at tagging calculated from the von Bertalanffy
growth function applied to length data from tagging events agreed with estimated
ages (t = -1.44, p = 0.16). Elapsed time between tagging and recapture varied from 6
to 10 years with a mean 8.4 ± 0.2 years. Mean growth was 11 ± 5 mm year-1 TL, with
an observed maximum and minimum growth rate of 19.3 mm year-1 and 4.6 mm year-
1, respectively (Figure 17). Annual growth rate was higher for larger fish (TL >800
mm, n = 10) than for smaller fish (TL <800 mm, n = 10), grouped by TL at time of
recapture (t = 2.1, p = 0.01) (Figure 18). Mean time at large was not significantly
different between size bins (t = 2.1, p = 0.4). Due to the absence of sufficient linear
correlation, ANCOVA was inappropriate (covariate = time at large) to analyze
differences in growth rate between size classes. Weight-based divisions of size at
tagging did not show growth rate differences.
Gear Selectivity
A visual inspection of the fitted Holt, Baranov and Regression models (Figure
19) was sufficient to recognize the Baranov model was the most representative for the
sample dataset. When R2 coefficients were calculated across mesh sizes (i.e., 10.2,
15.2, 17.8 cm), the Baranov model demonstrated the best fit (R2 = 0.53) to the
dataset, while the Regression (R2 = 0.3) and Holt (R2 = 0.28) model coefficients were
substantially lower. Standardizing selectivity curves to a common maxima of 1 was
necessary given the unknown size-structure of the population (Hamley 1975), though
it has been demonstrated through direct estimates of gear selectivity that gradations in
mesh size yield different capture efficiencies (Hamley & Reiger 1973).
Table 6. Growth during time at large for recaptures of previously marked shortnose sturgeon during field sampling.
WtTLFL Time at large
(years)Growth (mm)
Floy tag #PIT tag #Estimated
age (tagging)Predicted age
(tagging)Estimated age
(capture)
89
9.257.929.338.428.088.426.92
98
9.337.429.928.679.588.56.337.67
410A5400542270450803No Scan4138575F491F4875574D227D170F562223761752227D61401541455C147B223342765A413931453C1F485E0338410A5D414DNo TagNo Scan1F3D6B0732221E2C707F41132D5C1E2224160E70
15119151491879791212810661012
10181.912511511200.15872010190.088735067128160.9450486151.2159747151.531486716240.951301229170.381215511180.94100349180.4973513210.9340301120-0.114453057488150.5268654140.0460-1*048136150.68984007161 / 071627170.1646-13*03164120.51046516220.841164710180.424038
* Sturgeon showing negative growth (FL) during time at large.
WtTLFL Time at large
(years)Growth (mm)
Floy tag #PIT tag #Estimated
age (tagging)Predicted age
(tagging)Estimated age
(capture)
89
9.257.929.338.428.088.426.92
98
9.337.429.928.679.588.56.337.67
410A5400542270450803No Scan4138575F491F4875574D227D170F562223761752227D61401541455C147B223342765A413931453C1F485E0338410A5D414DNo TagNo Scan1F3D6B0732221E2C707F41132D5C1E2224160E70
15119151491879791212810661012
10181.912511511200.15872010190.088735067128160.9450486151.2159747151.531486716240.951301229170.381215511180.94100349180.4973513210.9340301120-0.114453057488150.5268654140.0460-1*048136150.68984007161 / 071627170.1646-13*03164120.51046516220.841164710180.424038
* Sturgeon showing negative growth (FL) during time at large.
0
20
40
60
80
100
120
140
160
180
0 2 4 6 8 10 12Time at large (years)
Gro
wth
(mm
)
19.3 mm year-1
4.6 mm year-1
0
20
40
60
80
100
120
140
160
180
0 2 4 6 8 10 12Time at large (years)
Gro
wth
(mm
)
19.3 mm year-1
4.6 mm year-1
Figure 17. Growth trajectories of tagged shortnose sturgeon recaptured during sampling. Mean growth rate during time at large was 11 ± 5mm year-1, with trajectories of growth exceeding the mean falling above the shaded area and those below falling within the shaded area. Maximum and minimum observed growth rates are indicated by the arrows.
58
y = 0.04 x - 22.3R2 = 0.50
0
5
10
15
20
25
600 700 800 900 1000TL at recapture (mm)
Incr
emen
t gro
wth
(mm
yr-
1)
y = 0.04 x - 22.3R2 = 0.50
0
5
10
15
20
25
600 700 800 900 1000TL at recapture (mm)
Incr
emen
t gro
wth
(mm
yr-
1)
Figure 18. Mean annual growth (mm yr-1) plotted against length at recapture (mm) for shortnose sturgeon tagged and recaptured from the Hudson River.
HG
ED F
I
B C
00.20.40.60.8
1
400 600 800 1000 500 700 900 1100 500 700 900 1100
500 700 900 1100500 700 900 11000
0.20.40.60.8
1
400 600 800 1000
Sele
ctiv
ity
(sta
ndar
dize
d to
1)
00.20.40.60.8
1
400 600 800 1000 500 700 900 1100
Total Length (mm)
500 700 900 1100
HG
ED F
I
B C
00.20.40.60.8
1
400 600 800 1000 500 700 900 1100 500 700 900 1100
500 700 900 1100500 700 900 11000
0.20.40.60.8
1
400 600 800 1000
Sele
ctiv
ity
(sta
ndar
dize
d to
1)
00.20.40.60.8
1
400 600 800 1000 500 700 900 1100
Total Length (mm)
500 700 900 1100
Figure 19. Selectivity curves for each mesh (10.2 cm mesh: open circles, 15.2 cm mesh: filled circles, 17.8 cm mesh: filled triangles) overlaid on catch values standardized to 1. Plots A-C are curves generated from the Holt model, plots D-F are curves generated from the Regression model, and plots G-I are generated from the Baranov model.
60
Each mesh size had a unique lO, that followed an intuitive pattern of
increasing lO with increasing mesh perimeter (Figure 20). Selectivity modifiers
yielded a falsely inflated catch adjustment for one fish that fell on the extreme upper
limb of the 15.2 cm mesh curve. This sturgeon measured 984 mm TL and was
estimated at 30 years old. Due to its large size and subsequent peripheral position on
the 15.2 cm selectivity curve, the adjusted catch was scaled up to 20 ‘fish’ (1900%
inflation). Following loge transformation, the adjusted catch value of 3.4 fell well
above three standard deviations (s.d. = 0.23) of the 30 year age-class catch of 0.06
predicted by the catch curve and was therefore considered a statistical outlier and not
included as a valid predictor of recruitment strength.
Recruitment
Catch data, adjusted for gear selectivity and effort, yielded an adjusted total
catch of Cadjust = 1238 sturgeon (Cactual = 554) (Table 7). Adjusted catch values
represent the number of sturgeon of each size-class that would have been caught if all
gear sizes were fished with equal effort and selectivity modifiers were identical across
sturgeon sizes. Adjusted catch-at-age peaked at 5 years and remained stable until 7
year before declining; therefore, fish of age ≥ 7 years were assumed to have fully
recruited to the gear and were included in the catch curve analysis. The maximum
age included in the catch curve was 23 years. Least-squares linear regression (R2 =
0.75) yielded a Z = -0.22 ± 0.03 (s.e.) (Figure 21 A). This implies that 20% of the
standing abundance of each cohort (age 5+) perishes annually, or a survivorship of
80% cohort-1 year-1. Based upon the protected status of the shortnose and assuming
0
0.2
0.4
0.6
0.8
1
500 600 700 800 900 1000 1100
Total Length (mm)
Sele
ctiv
ity
17.8 mesh ( )
10.2 mesh ( )
15.2 mesh ( )
lO = 700 mm
lO = 800 mm
lO = 825 mm
0
0.2
0.4
0.6
0.8
1
500 600 700 800 900 1000 1100
Total Length (mm)
Sele
ctiv
ity
17.8 mesh ( )
10.2 mesh ( )
15.2 mesh ( )
lO = 700 mm
lO = 800 mm
lO = 825 mm
Figure 20. Baranov selectivity curves generated for each of the mesh sizes. Optimal capture length lO is denoted by arrows for each mesh size.
62
Table 7. Catch by age class with mean Total Length (TL), catch per mesh, actual (Cactual) and adjusted (Cadjusted) catches of shortnose sturgeon over the course of field sampling. In the bottom row is estimated mortality, Z, calculated for each mesh, actual, and adjusted values, Z used in calculations, 0.22, given in boldface.
0.220.420.20.220.3-Z
30101030981
1101029830
6110026984
1101025865
4303023858
3312022851
4312021880
7615020858
9909019840
2726620018846
3131625017821
67641449116828
69651447415805
89761255914796
111951771713787
89621242812761
7840727611739
3424220210753
59140869678
106100648661
12750147593
12150146601
16580175561
17.815.210.2Σ CadjustedΣ Cactual
Mesh sizes (cm)Age, years
Mean TL, mm
0.220.420.20.220.3-Z
30101030981
1101029830
6110026984
1101025865
4303023858
3312022851
4312021880
7615020858
9909019840
2726620018846
3131625017821
67641449116828
69651447415805
89761255914796
111951771713787
89621242812761
7840727611739
3424220210753
59140869678
106100648661
12750147593
12150146601
16580175561
17.815.210.2Σ CadjustedΣ Cactual
Mesh sizes (cm)Age, years
Mean TL, mm
63
7 years
y = -0.22x + 6.8R 2 = 0.74
0
1
2
3
4
5
6
Lnca
tch
(adj
uste
d) 40 62
95
7665 64
31 26
1
9 6
3
1
1
1
33
2414
5 1058
Res
idua
ls o
f cat
ch c
urve
Age (years)
1999-1995
1994-1990
1989-1985
1984-1979
-3-2-10123
0 5 10 15 20 25 30
B
A
7 years
y = -0.22x + 6.8R 2 = 0.74
0
1
2
3
4
5
6
Lnca
tch
(adj
uste
d) 40 62
95
7665 64
31 26
1
9 6
3
1
1
1
33
2414
5 1058
Res
idua
ls o
f cat
ch c
urve
Age (years)
1999-1995
1994-1990
1989-1985
1984-1979
-3-2-10123
0 5 10 15 20 25 30
B
A
Figure 21. A) Natural log transformed adjusted catch (y-axis) plotted against age for Hudson River shortnose sturgeon. Least-squares linear regression was fitted to filled data given by filled circles only. Above each symbol is the associated subsample size. B) Catch curve residuals plotted against age (abscissa) with four year-class stanzas denoted by the dashed boxes.
64
negligible poaching/bycatch, the estimate of Z represents the natural mortality rate,
M.
Residuals from the catch curve described a slight, curvilinear pattern (Figure
21 B). Significant differences in residuals (1979-1999 cohorts) occurred between
age-classes when binned into 5-year stanzas (interval 1984-1979 included only 5 age-
classes though spanning 6 years) (Figure 21 B). Residuals of 5-9 & 20-26 year olds
were negatively biased in regard to predicted values and both showed a significant
difference from the 15-19 year old mean residual value (Table 8).
Observed recruitment strength index (RSI) values indicated that 1988 was the
strongest (RSI = 1.0) and 1979 the weakest year-class (RSI = 0.13) during the period
1979-1999 (Figure 22). Unadjusted age structure corroborated RSI results, with 13
year-olds comprising the most numerous age-class (n = 95). No recruits were
predicted from 1976, 1977 and 1980 due to a lack of capture data from the 24, 27 &
28 age classes. Hindcast estimates of yearling cohort abundances varied by up to an
order of magnitude among year-classes. The most abundant year-class was spawned
in 1988, which resulted in 42,659 yearlings while the most depauperate was the 1994
year-class, which yielded 4,886 yearlings (Figure 23).
Fall Juvenile Survey
Shortnose bycatch from the Hudson River Utilities Fall Juvenile Survey was
significantly correlated with predicted RSI values (Table 9). Iterative lagging of the
trawl data yielded progressively increasing correlation values from 2-6 years, though
significant correlations were present for each of the 4-8 year lag scenarios (rS > 0.46,
65
Table 8. Mean catch curve residual values grouped by age-class stanzas. See Figure 21 for catch curve. Means with any identical letters are not significantly different at α = 0.05 (Tukey-Kramer protected against experimentwise error inflation).
-0.6611 a1979-1984
0.4943 b1985-1989
0.2551 ab1994-1990
-0.6134 a1999-1995
Mean residual valueYear-class
21-26
16-20
11-15
5-10
Age-class
-0.6611 a1979-1984
0.4943 b1985-1989
0.2551 ab1994-1990
-0.6134 a1999-1995
Mean residual valueYear-class
21-26
16-20
11-15
5-10
Age-class
Mean year-class strength
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
1977
1975
Hatch year
Rec
ruitm
ent s
treng
th
inde
x (R
SI)
0
0.2
0.4
0.6
0.8
1
Mean year-class strength
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
1977
1975
Hatch year
Rec
ruitm
ent s
treng
th
inde
x (R
SI)
0
0.2
0.4
0.6
0.8
1
Figure 22. Index of yearly recruitment success based upon actual catch adjusted for gear selectivity, effort and cumulative mortality. Relative cohort strengths (y-axis) are plotted against hatch year (abscissa).
0
500010000
15000
2000025000
30000
3500040000
45000
19991997199519931991198919871985198319811979Year Class
Abundance
Minimum estimated age 1 cohort
abundance: 4,886
Maximum estimated age 1 cohort
abundance: 42,659
0
500010000
15000
2000025000
30000
3500040000
45000
19991997199519931991198919871985198319811979Year Class
Abundance
Minimum estimated age 1 cohort
abundance: 4,886
Maximum estimated age 1 cohort
abundance: 42,659
Figure 23. Estimated age 1 cohort abundance for Hudson River shortnose sturgeon, ages 5-30 years.
68
Table 9. Correlation of yearling recruitment strength index (RSI) values relative to shortnose sturgeon bycatch from the Hudson River utilities sponsored Juvenile Fall Survey, lagged over yearly increments.
0.040.528 years
0.620.149 years
0.020.567 years
0.010.586 years
0.050.485 years
0.050.464 years
0.190.333 years
0.85-0.052 years
0.12-0.341 year
0.02-0.6No lag
p-valueCorrelation (rS)Lag scenario
0.040.528 years
0.620.149 years
0.020.567 years
0.010.586 years
0.050.485 years
0.050.464 years
0.190.333 years
0.85-0.052 years
0.12-0.341 year
0.02-0.6No lag
p-valueCorrelation (rS)Lag scenario
69
p ≤ 0.05). Lagging trawl data by 6 years resulted in the highest correlation (rS = 0.58,
p = 0.01) (Figure 24).
Environment: Flow
Correlation analysis suggested significant correlation between RSI estimates
and mean monthly flow volumes (Figure 25). Flow was found to correlate with fall
months in the year preceding individual recruitment events (adult effect). November
flow (443.4 m3 sec-1) was positively correlated with RSI of yearlings spawned the
following year (rS = 0.51, p < 0.05), as was October flow (328.6 m3 sec-1, rS = 0.44, p
< 0.05). Mean fall flow volume (September, October, November) showed a similarly
significant correlation (327.7 m3 sec-1, rS = 0.53, p < 0.05). Correlations between
spring flow volume (May, June) and the RSI (egg and/or larval effect) were not
evident from the analysis. Analysis of the late juvenile period indicated a weak
correlation (rS = 0.43, p = 0.05) between yearling recruitment levels and mean fall
(September, October, November) flow volume (322.1 m3 sec-1).
Rec
ruitm
ent s
treng
th
(Rel
ativ
e va
lue)
Trawl survery
CPU
E (fish 1000 ft 3 -1)Trawl data lagged
6 years
Spearman rS = 0.58
Hatch year
2003
2001
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
1977
1975
0
0.2
0.4
0.6
0.8
1
0.01
0.02
0.03
0.04
0.05
0
0
0.2
0.4
0.6
0.8
1
0.01
0.02
0.03
0.04
0.05
0
Recruitment strength index
Trawl data
Rec
ruitm
ent s
treng
th
(Rel
ativ
e va
lue)
Trawl survery
CPU
E (fish 1000 ft 3 -1)Trawl data lagged
6 years
Spearman rS = 0.58
Trawl data lagged 6 years
Spearman rS = 0.58
Hatch year
2003
2001
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
1977
1975
0
0.2
0.4
0.6
0.8
1
0.01
0.02
0.03
0.04
0.05
0
0
0.2
0.4
0.6
0.8
1
0.01
0.02
0.03
0.04
0.05
0
Recruitment strength index
Trawl data
Figure 24. A) Hindcast annual recruitment strengths (open bars) versus shortnose sturgeon CPUE from the Hudson River utilities sponsored Fall Juvenile Trawl Survey. B) Trawl bycatch CPUE lagged by 6 years and overlaid on hindcast recruitment strengths.
Cohort strength (R
elative value)
Year-class strength
Flow
(m3
sec-1
) Flow volume
1974
1976
1980
1984
1986
1990
1994
1996
0
0.2
0.4
0.6
0.8
1
0
400
800
1200
1600
2000
1982
1988
1992
1978
Cohort strength (R
elative value)
Year-class strength
Flow
(m3
sec-1
) Flow volume
1974
1976
1980
1984
1986
1990
1994
1996
0
0.2
0.4
0.6
0.8
1
0
400
800
1200
1600
2000
1982
1988
1992
1978
Figure 25. Time series of mean monthly flow volume (ft3 minute-1, primary y-axis) collected from Green Island, New York (1946-2002, USGS) with hindcasted recruitment strength (secondary y-axis) overlaid on flow for corresponding years (Blow-up window) of 1974-1997.
72
Discussion
Recovery of shortnose sturgeon in the Hudson River coincided with a series of
strong year-classes from 1988-1991 but the demographic analysis did not support my
initial expectation of strong year-classes during the early 1980’s (Figure 23). Mean
recruitment levels (estimated as the subsequent year’s abundance of yearlings) from
1988-1991 was 36,331, compared to the preceding and succeeding periods of 1979-
1987 (mean = 15,361) and 1992-1999 (mean = 9,753). Yearling abundance
estimates varied 10-fold from 1979-1999, reflecting episodic recruitment success
often associated with a ‘periodic’ reproductive strategy (Winemiller & Rose 1992),
which has been associated with long-lived, highly fecund species such as Acipenser
(Bain 1997, Bemis & Kynard 1997, Secor & Waldman 1999). Counter to the
hypothesized response, year-class abundance showed no significant correlations with
the magnitude of annual spring freshets, possibly due to a non-linear response and/or
the exclusion in the analysis of water temperature, which may act in concert or
confound the influence of flow. It was not possible to analyze recruitment success
concomitant with the return of system normoxia; yet retrospective analysis of year-
class strength and growth rates based upon my demographic analysis provided
circumstantial evidence (e.g., presence of strong year-classes) for improved
recruitment, growth rates, and physical condition (no observed malformation or fin
rot) during the recent period of system recovery to summertime normoxia.
As a prerequisite to determining population demographics, annuli in pectoral
spines were validated as an appropriate ageing structure for Hudson River shortnose
sturgeon. The technique yielded age estimates that were found to be non-biased
73
across ages and over time in repeated trials. The yearly formation of annuli was
validated through Marginal Increment Analysis (MIA). The timing of annuli
formation typically occurred during the over-wintering period (November-March);
presumably a period of low metabolic activity and reduced growth. Growth rate (TL)
for the Hudson River population was more rapid than size-at-age data from the late
1970’s suggest (Dadswell 1984), though asymptotic size appears to have declined
slightly since that time. The recent increase in growth rate (i.e., compared to 1980
population) may be indicative of improvements in ambient water quality mediated by
density-dependent factors that in more recent years could be limiting growth rates.
Growth rate was intermediate relative to populations at the extremes of the species’
range and was similar to those systems in close latitudinal proximity to the Hudson
River. The Hudson River population appears to experience a higher than expected
annual mortality of 20% yr-1 (Z = -0.22), though this estimate is predicated on the
assumption of constant recruitment and is likely inflated relative to the actual annual
mortality rate.
Age Determination
Age estimates of wild shortnose sturgeon based upon annuli observed in
pectoral fin spines were relatively precise, with precision index values comparing
favorably to studies of other Acipenserids and similar moderate to long-lived species
(Table 10). Lack of an age bias in CV values indicated that shortnose fin spines can
provide precise ages for individuals 5-30 years old. The discrepancy in CV and APE
in the literature may be attributed to the disparity in estimated age maxima for each
species (Campana 2000). At older ages, CV and APE values are expected to rise due
74
Table 10. Precision in ageing studies of Acipenser and other fish species of varying longevity, reported as average percent error (APE) and coefficient of variation (CV).
18.534.0 - 4.2-European eel Anguilla anguilla e
30-10.4Goldband snapper Pristipomoides mutidens c
157.5 - 7.9-Common carp Cyprinus carpio g
111.21.2Northern pike Esox luscius h
-
-
6.9 - 12.7
3.0
-
5.9
APE
116.3Monkfish Lophius vomerinus i
304.0Shortnose sturgeon A. brevirostrum
20-Starspotted smoothound Mustelus manazo d
162.8Thorny skate Amblyraja radiata f
Acipenseridae
424.8Atlantic sturgeon A. oxyrinchus b
1047.8White sturgeon A. transmontanus a
Maximum estimated ageCVSpecies
a - h Source: (a) Rien & Beamesderfer (1994); (b) Stevenson & Secor (1999); (c) Newman & Dunk (2003); (d) Cailliet et al. (1990); (e) Svedang et al. (1998); (f) Sulikowski et al. (2005); (g) Vilizzi & Walker (1999); (h) Laine et al. (1991); (i) Maartens et al. (1999)
18.534.0 - 4.2-European eel Anguilla anguilla e
30-10.4Goldband snapper Pristipomoides mutidens c
157.5 - 7.9-Common carp Cyprinus carpio g
111.21.2Northern pike Esox luscius h
-
-
6.9 - 12.7
3.0
-
5.9
APE
116.3Monkfish Lophius vomerinus i
304.0Shortnose sturgeon A. brevirostrum
20-Starspotted smoothound Mustelus manazo d
162.8Thorny skate Amblyraja radiata f
Acipenseridae
424.8Atlantic sturgeon A. oxyrinchus b
1047.8White sturgeon A. transmontanus a
Maximum estimated ageCVSpecies
a - h Source: (a) Rien & Beamesderfer (1994); (b) Stevenson & Secor (1999); (c) Newman & Dunk (2003); (d) Cailliet et al. (1990); (e) Svedang et al. (1998); (f) Sulikowski et al. (2005); (g) Vilizzi & Walker (1999); (h) Laine et al. (1991); (i) Maartens et al. (1999)
75
to increased difficulty in interpreting narrow annuli associated with decreased growth
rates as the fish approaches asymptotic size (Casselman 1987, Campana et al. 1995).
Evidence supported seasonal elaboration of an annulus by Hudson River
shortnose sturgeon over an annual cycle. Marginal increment analysis indicated that
annuli were fully formed sometime between the late fall and early spring months.
This timing of annulus formation agrees with a similar study by Stevenson and Secor
(1999) on a sympatric population of Atlantic sturgeon. Ideally, here and elsewhere,
annulus formation should be verified across many age classes (Campana 2000).
Because the shortnose sturgeon is listed as an Endangered Species, we were unable to
target a sufficiently large sample to undertake age-class specific marginal increment
analysis. Thus, our marginal increment analysis was undertaken for pooled year-
classes, which limits its generality across all ages.
I was unable to accurately estimate the known age of hatchery-raised
shortnose sturgeon. Errors in age determination of captive sturgeons (e.g., white
sturgeon, Brennan & Cailliet 1991; pallid sturgeon, Hurley et al. 2004) and of reared
individuals of other species (e.g., Atlantic herring Clupea harengus, Lough et al.
1980; starry flounder Platichthys stellatus, Campana 1984; walleye Sander vitreus,
Parrish et al. 1994), is well-documented and has been associated with artificial
conditions experienced by the fish within the rearing or holding facility. In the wild,
reinforced annual cycles related to light, temperature, salinity, feeding, and
reproduction are probably critical to annulus timing; these cycles are not well
simulated in most hatchery environments.
76
Limits to precision and accuracy in interpreting annuli in fin spines in
hatchery-raised fish included supernumary annuli, bands of narrow annuli, variable
widths of annuli, inclusion of embedded rays within the primary spine, resorption or
deposition of calcareous material and physical deterioration. Supernumary annuli
were especially prevalent across the first three to five annuli of estimated age and
may constitute a source of age overestimation if not correctly or consistently
interpreted. Variable spacing of annuli has been hypothesized to stem from
endogenous and exogenous factors that cause fish to grow slowly (Roussow 1957,
Rossiter et al. 1995). The reduction in somatic growth may be recorded in the
relative widths of annuli. Spawning bands are thought to chronicle a reduction in
somatic growth as energy is shunted to gonadal development over a period of several
years prior to a spawning event (Roussow 1957). If annuli are not clearly contrasted
or are tightly banded, ages are likely underestimated.
Age validation studies have been reported for Siberian sturgeon (A. baerii)
(Sokolov & Akimova 1976), white sturgeon (Brennan & Cailliet 1991, Rien &
Beamesderfer 1994, Paragamian & Beamesderfer 2003), lake sturgeon (A. fulvescens)
(Rossiter et al. 1995), Atlantic sturgeon (A. oxyrinchus) (Stevenson 1997, Stevenson
& Secor 1999), pallid sturgeon (Scaphirhynchus albus) (Hurley et al. 2004), and
shovelnose sturgeon (S. platorynchus) (Whiteman et al. 2004). These studies have
used a suite of validation techniques, including mark-recapture experiments with
oxytetracycline markers and traditional tagging or a combination of the two (Brennan
& Cailliet 1991, Rien & Beamesderfer 1994, Rossiter et al. 1995, Stevenson & Secor
1999, Paragamian & Beamesderfer 2003), marginal increment analysis (Stevenson &
77
Secor 1999, Whiteman et al. 2004), microchemical analysis (e.g., calcium
concentrations) (Stevenson & Secor 1999), radiometric ageing (Burton et al. 1999),
and the use of known-age individuals (Sokolov & Akimova 1976, Brennan & Cailliet
1991, Stevenson & Secor 1999, Hurley et al. 2004). Attempts to validate age
estimates have yielded mixed results, with few providing compelling evidence
(Sokolov & Akimova 1976, Rossiter et al. 1995). Some studies support partial
validation over a limited range of age-classes (Brennan & Cailliet 1991, Stevenson &
Secor 1999; this study) and the majority of studies conclude that annulus formation
was temporally inconsistent and/or annuli were difficult to interpret (Rien &
Beamesderfer 1994, Burton et al. 1999, Paragamian & Beamesderfer 2003, Hurley et
al. 2004, Whiteman et al. 2004). These problems underscore the importance of
validating the relationship between counts of annuli and known age for each species.
Growth
Age estimates indicate high intra-cohort variability in size for the Hudson
River population, a characteristic common to Acipenser spp. (Kolhorst et al. 1980,
Nakamoto et al. 1995, Stevenson & Secor 1999). Sexual dimorphism of shortnose
sturgeon probably constitutes a substantial portion of the observed variability in size-
at-age. In a review article by Greeley (1937), age and length data are given for male
(n=34) and female (n=47) shortnose sturgeon captured from the Hudson River. The
largest male was 713 mm TL and 7 years of age, while the largest female was 883
mm TL and 13 years of age. Mean length of males was greater than females for ages
5 (♂: 560 mm, ♀: 513 mm) and 7 (♂: 652 mm, ♀: 605 mm), indicating more rapid
growth of males during the sub-adult stage. In the St. John River, Dadswell (1979)
78
observed similar sex-specific growth attributes, with males experiencing higher initial
growth rates and attaining a lower maximum length (K = 0.063, FL∞ = 1087 mm)
than females (K = 0.047, FL∞ = 1270 mm).
Recapture of tagged Hudson River shortnose sturgeon provided a second
source of growth estimates that generally corroborated my age estimates and growth
models. Still, without known ages for the marked fish, the accuracy of predicted and
estimated ages remains uncertain. Mark-recapture studies of sturgeon spp. have
yielded disparate results; an analysis of white sturgeon mark-recapture data from the
Kootenai River, ID showed that age estimates derived from fin-spines significantly
underestimated true ages (Paramagian & Beamesderfer 2003). Interestingly, in my
study there was a significant difference in the yearly growth increment (mm yr-1)
between larger and smaller sturgeon with smaller sturgeon at tagging experiencing
lower growth rates than larger fish. This result runs counter to the expectation that
smaller sturgeon would demonstrate higher mean growth yr-1 relative to older fish as
has been reported in tag-effect studies on other fishes (e.g., northern pike Scheirer &
Coble 1991, Arctic grayling Thymallus arcticus Hughes 1998, tropical goby
Coryphopterus glaucofraenum Malone et al. 1999). This discrepancy could indicate
that smaller sturgeon were disproportionately stressed by handling and tagging and
therefore experienced significant post-capture growth depression. Alternatively,
those fish that were larger at the time of tagging may have been predominantly
females. The smaller "males" would be constrained by a smaller asymptotic size and
evince lower growth. Sexual dimorphism on this scale does not seem unreasonable.
Greeley (1937) reported a maximum size for male Hudson River shortnose sturgeon
79
of 713 mm which is well below the mean tagging TL of 772 mm for the larger size
bin. This interpretation is based upon the assumption that the smaller fish at tagging
were predominately males and not simply younger females, an assumption that
remains speculative due to lack of data.
Growth parameters L∞ and K were intermediate in relation to the set of values
estimated for shortnose sturgeon across their range (Table 11). Shortnose sturgeon
populations tend to exhibit, more rapid growth (Kmax = 0.149) and a smaller
maximum size (FL∞ min = 870 mm) in southern estuaries compared to northern
systems (Kmin = 0.042 & FL∞ max = 1300 mm; Figure 26; Dadswell et al. 1984). The
Hudson River FL∞ (925 mm) was similar to literature values for the proximal
Connecticut (1000 mm) and Kennebec (938 mm) Rivers.
Growth parameter values from a previous study of the same Hudson River
population reported TL∞ = 1234 mm (calculated from 1064 mm FL) and K= 0.044
(Dadswell et al. 1984), which suggests a temporal shift in growth dynamics. In
general, shortnose sturgeon are currently attaining a larger size at age than they did
20-40 years ago (Figure 27). Factors that would prompt an increase in growth
include improved water quality, increased nursery habitat, and conditions favoring
greater forage.
A recent increase in specific growth rates of Hudson River shortnose sturgeon,
particularly during the juvenile stage when habitats are limited, may reflect the return
of large portions of the river to year-round normoxia. Hypoxia can limit the function
of most bioenergetic processes, resulting in reduced growth and/or increased
mortality (Fry 1971, Niklitschek & Secor 2005). Regulations promulgated under the
80
Table 11. Growth parameters from the von Bertalanffy growth function estimated for shortnose sturgeon across their range. Boldface row indicates parameters derived during this study.
0.08-0.12
0.22
0.12
0.12-0.15
Z
Current study-7.050.064925Hudson R., NY
Combined
Altamaha R., GACombined
Pee Dee-Winyah, SCFemaleMaleCombined
Hudson R., NYFemaleMaleCombined
Connecticut. R.(Lower river, CT)
Combined
Connecticut R. (Holyoke Pool, MA)
Combined
Kennebec R., MECombined
Saint John R., FLFemaleMaleCombined
River System
*Heidt & Gilbert, 1978-3.150.14997
*Marchette & Smiley, 1982
-2.33-4.5
-6.02
0.133 0.114 0.093
83.873.987
*Greely, 1937
*Dovel, 1981
-3.17-1.86.39
0.0790.3050.044
102.657.9
106.4
*Buckley, unpublished-2.730.073100
*Taubert, 1980b-2.640.08487.8
*Squiers & Smith, 1978 -3.890.09893.8
Dadswell, 1979-1.10.79-1.96
0.0470.0630.042
127108.7130
*SourcetKFL∞
0.08-0.12
0.22
0.12
0.12-0.15
Z
Current study-7.050.064925Hudson R., NY
Combined
Altamaha R., GACombined
Pee Dee-Winyah, SCFemaleMaleCombined
Hudson R., NYFemaleMaleCombined
Connecticut. R.(Lower river, CT)
Combined
Connecticut R. (Holyoke Pool, MA)
Combined
Kennebec R., MECombined
Saint John R., FLFemaleMaleCombined
River System
*Heidt & Gilbert, 1978-3.150.14997
*Marchette & Smiley, 1982
-2.33-4.5
-6.02
0.133 0.114 0.093
83.873.987
*Greely, 1937
*Dovel, 1981
-3.17-1.86.39
0.0790.3050.044
102.657.9
106.4
*Buckley, unpublished-2.730.073100
*Taubert, 1980b-2.640.08487.8
*Squiers & Smith, 1978 -3.890.09893.8
Dadswell, 1979-1.10.79-1.96
0.0470.0630.042
127108.7130
*SourcetKFL∞
* Growth parameters as given by Dadswell et al. 1984
81
0
200
400
600
800
1000
1200
1400
30 34 38 42 46
Latitude (O North)
FL ∞
(mm
)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
Grow
th coefficient: kFL ∞
k
0
200
400
600
800
1000
1200
1400
30 34 38 42 46
Latitude (O North)
FL ∞
(mm
)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
Grow
th coefficient: kFL ∞
k
Figure 26. Von Bertalanffy growth parameters FL ∞ (primary y-axis) and k (secondary y-axis) plotted against latitude.
82
1 Pee Dee-Winyah R.
2 Hudson R.
3 Hudson R.
4 St. John R.
FL∞ = 870, k = 0.093
FL∞ = 940, k = 0.10
FL∞ = 1060, k = 0.044
FL∞ = 1300, k = 0.042
400
600
800
1000
0 5 10 15 20 25 30 35Age (years)
Fork
Len
gth
(mm
)
500
700
900
1 Pee Dee-Winyah R.
2 Hudson R.
3 Hudson R.
4 St. John R.
FL∞ = 870, k = 0.093
FL∞ = 940, k = 0.10
FL∞ = 1060, k = 0.044
FL∞ = 1300, k = 0.042
1 Pee Dee-Winyah R.
2 Hudson R.
3 Hudson R.
4 St. John R.
FL∞ = 870, k = 0.093
FL∞ = 940, k = 0.10
1 Pee Dee-Winyah R.
2 Hudson R.
3 Hudson R.
4 St. John R.
FL∞ = 870, k = 0.093
FL∞ = 940, k = 0.10
FL∞ = 1060, k = 0.044
FL∞ = 1300, k = 0.042
400
600
800
1000
0 5 10 15 20 25 30 35Age (years)
Fork
Len
gth
(mm
)
500
700
900
Figure 27. Predictions of Fork Length-at-age from von Bertalanffy growth models (combined sex models) for three different estuaries across the latitudinal range of shortnose sturgeon. Sources for each curve: 1) Dadswell et al. 1984, 2) current study, 3) Dadswell et al. 1984, 4) Dadswell 1979.
83
Clean Water Act (1970) led to increased summertime DO levels across a substantial
portion of the Hudson River Estuary (Leslie et al. 1988), from Troy’s Federal Dam
extending c. 60 km downstream. This area constitutes c. 40% of the larval/juvenile
nursery habitat for shortnose sturgeon as well as summer foraging habitat used by
adult fish (Dovel 1980, Bain 1995). Laboratory studies have shown that YOY and
yearling shortnose sturgeon are unusually sensitive to hypoxia and demonstrate
decreased routine metabolism, consumption, feeding metabolism, growth and
survival at DO levels < 40% saturation (Jenkins et al. 1994, Niklitschek 2001, Secor
& Niklitschek 2001, Campbell & Goodman 2004).
Older juveniles and sub-adults may experience secondary effects of hypoxia if
it limits their distribution to smaller or less profitable foraging and refuge habitats.
The cascading effects of seasonal hypoxia can substantially reduce the biomass of
benthic macrofauna and diminish an ecosystem’s ability to transfer energy to higher
trophic levels in estuarine systems (Baird et al. 2004, Eby et al. 2005). Invertebrate
benthic macrofauna constitute the principal prey source of the shortnose sturgeon diet
(Dadswell 1979, Carlson & Simpson 1987). With a reduction in chronic seasonal
hypoxia, the sturgeon population would experience a simultaneous increase in
suitable foraging habitat during the important summer growth period and a
concomitant rise in available forage densities. Both of these factors could contribute
to the higher growth rates observed for the Hudson River population.
Exposure and uptake of PCBs can initiate an array of physiological responses
including disruption of the endocrine system and subsequently irregular growth and
development. In conjunction with increased summertime DO levels, reduced
84
contaminant concentrations may explain increases in sturgeon growth rates during
recent decades. Time series data show a >95% decline in PCB body burden for white
perch (Marone americana,31→3 ppm) and American eel (Anguilla rostrata, 77→2
ppm) in the lower Hudson River from 1978-1996 (NYSDEC 2001), two species
whose habitats overlap that of shortnose sturgeon. Recent body burdens appear to
have stabilized near 3 ppm among sampled fishes (e.g., striped bass M. saxatilis,
American shad, largemouth bass Micropterus salmoides, channel catfish Ictalurus
punctatus) below the Federal Dam (NYSDEC 2001).
Sublethal levels of bioaccumulated PCB’s have been found to retard growth in
channel catfish (Ictalurus punctatus, Hansen et al. 1976), larval Atlantic croaker
(Micropogonias undulatus, McCarthy et al. 2003) and brook trout (Salvelinus
fontinalis, Ndayibagira et al. 1995) and may manifest a similar effect in shortnose
sturgeon as has been proposed for white sturgeon (Kolhorst et al. 1980).
Developmental malformations and disease have been documented for PCB-exposed
sturgeon populations. Doyon et al. (1999) found that PCB induction of an enzyme
(cytochrome P-450) involved in the metabolism of a biologically active form of
vitamin A may be associated with limb and craniofacial anomalies in St. Lawrence
River lake sturgeon. In Hudson River shortnose sturgeon samples collected from
1975-1980, Dovel et al. (1992) reported the presence of a facial abnormality termed
“u-snout” in c. 2% of the captured sturgeon; a similar facial abnormality to that noted
among lake sturgeon sampled from the St. Lawrence River near Montreal (Doyon et
al. 1999). No incidents of facial malformation was observed during our field
collections (n = 596).
85
Recruitment
Generating a valid estimate of population age structure and index of
recruitment strengths required adjustment of the original catch values. Actual catch
(Cact = 597) was scaled to adjusted catch (Cadj = 1238) following compensation for
size-selectivity and effort. This procedure was similar to that used by Dadswell
(1979) for the St. John River population. Subsequent estimates of annual mortality,
age structure and recruitment were based on the adjusted catch.
Gear Selectivity
Application of the Baranov gear selectivity model involves several
assumptions, the most important of which is the principle of geometric similarity:
selectivity curves for different mesh size are similar because the ratio of mesh and
fish profile is similar across mesh sizes (Hamley 1975). One of the corollaries
implicit in the principle of geometric similarity is that the ratio of mesh size to
optimal capture length, K, is equal across mesh sizes (Hamley 1975). Results from
field sampling did not support this corollary; calculated values of K increased
progressively from 0.15 to 0.22 for the 10.2 and 17.8 cm mesh nets. These values are
often interpreted in terms of body morphology: thin bodied fishes (e.g., northern pike,
mackerel Scomber spp.) exhibit K values of 0.10-0.15, while deep body fishes (e.g.,
sunfishes Lepomis spp.) typically have values > 0.2 (Hamley 1975). The observed
increase in K corresponds with a positive shift in the girth to length ratio with
increased age of sturgeon, a general ontogenetic change observed in many fishes.
Selectivities of different mesh sizes are not necessarily equal in terms of catch
efficiency for a single species. Hamley & Regier (1973), using direct estimates of gill
86
net selectivity on walleye (S. vitreus), showed increased overall selectivity in larger
mesh sizes. The same pattern of increasing catch frequency with increased mesh size
was observed by Dadswell (1979) for the St. John River population of shortnose
sturgeon. In my study, this bias would tend to amplify observed patterns of stronger
year-classes historically than in recent times. Although indirect methods such as
those employed here are not preferred and may introduce bias, studies continue to
employ indirect methods of estimating gear selectivity (see Milton et al. 1998,
Anderson & Neumann 2000) as a practical and necessary alternative to direct
methods.
Mortality
The total instantaneous mortality rate Z = 0.22 for Hudson River shortnose
sturgeon was somewhat higher than that reported for other systems. A theoretical
estimate of natural mortality (M) based on observed longevity (tmax = 30) yielded M =
0.15 (Hoenig 1983). Estimates of total mortality may be viewed as a proxy for M as
there is no extant fishery for shortnose sturgeon anywhere in the United States. In the
most synoptic demographic shortnose sturgeon study to date, Dadswell (1979)
estimated Z = 0.12, 0.15 (1974, 1975) for ages 15-55 years from the Saint John River
(Table 11). Dadswell’s estimate included a fishing mortality component F (0.07)
stemming from bycatch in commercial fisheries for other species. Although a limited
American shad (Alosa sapidissima) fishery still exists on the Hudson River in which
shortnose have been observed as incidental bycatch (S. Nack, pers. com.), the fishery
is small and net deployment time is brief; therefore, shortnose sturgeon bycatch
mortality is likely to be negligible. The Hudson River population is considered the
87
largest (NMFS 1998) and as such could be exhibiting density-dependent mortality not
apparent in other systems where shortnose sturgeon occur. Additionally, the catch-
curve analysis assumes constant recruitment, yet I have shown that recruitments have
substantially varied over the past 25 years. For the presented catch curve (Figure 21
A), higher residuals for ages 12 to 17 and negative residuals for ages >17 would result
in a shift in slope towards a more negative value than were residuals more
homogenously distributed. An overestimation of Z would tend to underestimate of
early year-class strengths and overestimate more recent year-class strengths.
Age-structure
Age-structure for unexploited, long-lived fish populations is typified by a
wide range of successive age-classes composed of progressively fewer individuals
(Gill & Weatherley 1987; Secor 2000). Although the distribution of ages was likely
influenced by gear selectivity, this pattern agrees with the trend in demographic
structure observed for the Hudson population. Juveniles younger than 5 years of age
were not captured; reflecting the likely escape of smaller sturgeon from the sampling
gear. Also, juveniles do not tend to display the strong aggregating behavior of the
adults (Bain 1997, Haley 1999). The oldest sturgeon reported for the Hudson River
was 37 years old (Dadswell et al. 1984) and it seems likely that individuals of that age
still exist within the population at abundances below the detection level of this study
(i.e., a sampling rate of c. 1%).
The recruitment strength index (RSI) highlights a trend in increased
recruitment success that occurred 11-18 years ago (1993-1986). This period of
enhanced recruitment is corroborated through bycatch of adult shortnose sturgeon
88
during the Hudson River Utilities Fall Juvenile Survey (Figure 24). Catch rates from
the survey showed a temporal pattern of fluctuation similar to that of our hindcast
recruitment index. Because survey gear and sampling design are standardized, inter-
annual changes in the trawl survey shortnose CPUE should accurately reflect changes
in annual abundance of those fish vulnerable to the gear. Mean size at capture in the
trawl survey was 670 mm TL (Bain et al. 1997), which would correspond to 6-8 years
old (TL ~ 650 mm; Figure 16). This suggests a lag of c. 6 years between the
formation of strong year classes and increased relative abundance in the trawl survey,
a prediction that was substantiated through correlation analysis (Table 9).
Predictions of yearling abundances across year-classes indicated that
recruitment of Hudson River shortnose sturgeon varied by an order of magnitude.
Year-class variability of this magnitude is common among moderately to long-lived
estuarine and coastal fishes and reflects a “periodic” life-history strategy that depends
on conditions favorable for survival and growth of early life-stages that occur at an
interval less than the average life-span of the organism (Winnemiller & Rose 1992;
Warner and Chesson 1995). In the case of the Hudson River shortnose sturgeon, I
observed two c. 5-year periods of relatively low recruitments bridged by a 4-year
period of high recruitment (Figure 22). Interestingly, in more recent years lower
recruitments are coincident with record high abundance levels. Should recruitments
become density dependent as the population approaches carrying capacity, variability
in year-class strength might be expected to dampen as regulation becomes a more
dominant force.
89
Water pollution is listed as a principal factor in the decline of shortnose
populations (NMFS 1998). The return of normoxia to a substantial portion of the
tidal freshwater portion (c. 60 km) of the Hudson River Estuary (Leslie et al. 1988)
during the 1970’s occurred in a critical habitat for shortnose sturgeon. The location
of the former hypoxic zone, adjacent and downstream of the spawning grounds,
would have been particularly detrimental to larvae and juveniles. Mortality due to
hypoxic-anoxic conditions may result from direct mortality, reduced production
resulting from increased metabolic costs (Secor and Niklitschek 2001), and
synergistic interactions among stressors (i.e., a “habitat squeeze” sensu Coutant 1985,
Niklitschek & Secor 2005). Therefore, the return of normoxia during a crucial period
of growth and development (i.e., summer months) may have eliminated a substantial
recruitment bottleneck to the Hudson River population. Slow recovery of the benthic
ecosystem could explain the observed lag of c. 10 years (1978-1988) between system
return to normoxia and the first measurable strong yearling recruitment.
There is evidence that annual reproduction relies on an environmentally-
mediated endogenous mechanism in some fishes (de Vlaming 1972, Buckley &
Kynard 1985). The positive correlation of fall flow volume to recruitment success of
Hudson River shortnose sturgeon the following year may indicate that flow acts as a
primary component of a suite of environmental cues that initiate the final stages of
gonadal development. Alternatively, flow may be acting as an indicator of water
temperature or other environmental parameter exercising a more direct effect on
conditioning of adult sturgeon. Pre-spawn conditioning has been recognized as an
important component of inducing gametogenesis in captive white sturgeon (Webb et
90
al. 1999, Webb et al. 2001). Ambient water temperature during the gametogenic and
vitellogenic stages has been shown to be a significant factor in determining the
success of subsequent spawning of white sturgeon (Doroshov et al. 1997, Webb et al.
1999). These results support further study into the effects of fall flow on pre-spawn
conditioning of ovigerous shortnose sturgeon.
Although spring flow has frequently been implicated in recruitment of
anadromous fishes (Creco & Savoy 1984, Maurice et al. 1987, Stevens et al. 1987,
Paragamian & Wakkinen 2002), there was no correlation between spring flows and
year-class strength. This may stem from a non-linear response of spawning activity
to ambient flow rate on the spawning beds (Veshchev 1982, Buckley & Kynard 1985,
Kynard 1997). Moderate to swift current velocities (0.4-1.8 m sec-1) appear to trigger
spawning activity (Dadswell et al. 1984, Kynard 1997, but see Duncan et al. 2004),
while periods of extremely high discharge have been shown to deter spawning
(Buckley & Kynard 1985). Also, spring flow is likely to be confounded with
temperature effects since shortnose spawning tends to occur within a narrow
temperature window, ranging from 9-16oC (Taubert 1980, Dadswell et al. 1984,
Buckley & Kynard 1985, Duncan et al. 2004). Further correlative analyses using a
long-term data set of water temperature near the spawning grounds may help
elucidate the relationship between environmental variables and annual recruitment
success.
Management Implications
The shortnose sturgeon is showing signs of strong recovery in the Hudson
River although many population segments, especially in the south, still display low
91
abundance (or a lack of formal demographic assessment to make such a
determination). The Hudson River population appears to satisfy the formal delisting
criterion as defined by the Shortnose Sturgeon Recovery Team in which sufficient
abundance exists to 1) defray extinction risk and 2) make the loss of genetic diversity
unlikely (NMFS 1998). In addition, this study has shown that the population is
composed of a multi-tiered age structure, displays expected growth characteristics for
the species and population, and has experienced substantial recruitment (> 4,500
yearlings) annually for over two decades. Thus, the Hudson River population
probably represents an Endangered Species success story, the first of its kind among
listed fishes. Though a formal delisting process following recovery does not exist,
this population could serve as a template for evaluation and future delisting of other
population segments or species. Delisting shortnose sturgeon in the Hudson River
could expand opportunities for scientific study of the species, the findings of which
could prove beneficial to the species as a whole. Such research could include
ecological studies of YOY life stages, telemetry tracking of juvenile-adults,
quantification of vital rates across life stanzas, and environmental influences on
behavior.
The Hudson River shortnose sturgeon population also offers a unique
opportunity to employ physical (e.g., PIT, Floy ®, Carlin tags) or bio-assimilated
labels (i.e., Oxytetracycline) as methods for further validating the yearly formation of
annuli. Shortnose sturgeon of the Hudson River are optimal candidates for these
procedures because the robust demographic structure of the population equates to
ontogenetic stanzas of juveniles to mature adults available for validation of annulus
92
formation. Capture, marking, release and subsequent resampling requires direct
contact between researchers and sturgeon, a scenario with the potential to inflict high
stress levels on an Endangered Species. However, the Hudson River supports the
largest population of shortnose sturgeon in terms of numerical abundance (Bain 2001)
and therefore any stress-induced mortality would likely have trivial effects on this
population relative to other, less abundant populations. Use of Hudson River
shortnose as brood-stock to produce tagged known-age sturgeon to be released to the
river would serve as another powerful method of age validation in this species. Yet,
care would be required to avoid genetic swamping and reductions to the effective
population size due to the use on insufficient brood numbers.
93
Appendix
Shortnose sturgeon sampling data collected from Hudson River, November 2003 – November 2004. Sampling locations (Locale): Esopus Meadows area (Esopus), Catskill-Burden Dock area (Burden), and Albany area (Albany). Except for Clip # 299 (juvenile Atlantic sturgeon), all data entries refer to shortnose sturgeon.
Clip # Age FL
(mm) TL
(mm) Weight
(kg) Mesh size
(cm) Sample event Season Locale
1 17 652 737 2.3 15.2 1st fall Esopus 2 16 680 771 2.8 15.2 1st fall Esopus 3 15 721 840 4.5 15.2 1st fall Esopus 4 11 598 680 1.65 15.2 1st fall Esopus 6 16 683 800 3 15.2 1st fall Esopus 7 15 599 699 1.6 15.2 1st fall Esopus 8 13 575 646 1.5 15.2 1st fall Esopus 9 9 628 736 2.1 15.2 1st fall Esopus 10 12 686 755 2.9 15.2 1st fall Esopus 11 19 636 736 2 15.2 1st fall Esopus 12 19 721 839 3.2 15.2 1st fall Esopus 13 17 642 751 2.2 15.2 1st fall Esopus 14 14 751 865 3.6 15.2 1st fall Esopus 15 9 635 715 2.1 15.2 1st fall Esopus 16 15 655 745 2.1 15.2 1st fall Esopus 17 15 706 820 2.9 15.2 1st fall Esopus 18 16 632 741 2.4 15.2 1st fall Esopus 19 14 602 686 1.8 15.2 1st fall Esopus 20 15 682 788 2.7 15.2 1st fall Esopus 21 16 632 729 2.5 15.2 1st fall Esopus 22 11 552 640 1.5 15.2 1st fall Esopus 23 18 662 772 3.1 15.2 1st fall Esopus 24 15 680 771 2 15.2 1st fall Esopus 25 22 576 672 2 15.2 1st fall Esopus 26 15 551 749 2.9 15.2 1st fall Esopus 27 14 638 734 2.3 15.2 1st fall Esopus 28 13 610 710 2 15.2 1st fall Esopus 29 14 649 748 2.6 15.2 1st fall Esopus 30 17 621 717 2.5 15.2 1st fall Esopus 31 9 665 782 3 15.2 1st fall Esopus 32 16 580 700 2.3 15.2 1st fall Esopus 33 16 667 775 3 15.2 1st fall Esopus 34 15 706 820 2.9 15.2 1st fall Esopus 35 14 630 746 2.7 15.2 1st fall Esopus
94
36 14 555 655 1.8 15.2 1st fall Esopus 37 15 611 720 2.9 15.2 1st fall Esopus 38 16 758 883 5 15.2 1st fall Esopus 39 13 666 766 2.1 15.2 1st fall Esopus 40 13 662 765 2.9 15.2 1st fall Esopus 41 20 711 848 3.3 15.2 1st fall Esopus 42 13 723 862 3.3 15.2 1st fall Esopus 43 12 621 722 2 15.2 1st fall Esopus 44 5 609 724 2.5 15.2 1st fall Esopus 45 15 513 607 1.2 15.2 1st fall Esopus 46 17 750 866 4.3 15.2 1st fall Esopus 47 11 803 930 4.8 15.2 1st fall Esopus 48 15 625 708 2.2 15.2 1st fall Esopus 49 20 602 705 2.3 15.2 1st fall Esopus 50 13 681 774 3.1 15.2 1st fall Esopus 51 13 579 676 2.2 15.2 1st fall Esopus 52 14 679 786 3.1 15.2 1st fall Esopus 53 11 688 785 2.6 15.2 1st fall Esopus 54 17 575 653 1.8 15.2 1st fall Esopus 55 17 720 823 3.1 15.2 1st fall Esopus 56 13 692 782 2.6 15.2 1st fall Esopus 57 14 640 757 2.4 15.2 1st fall Esopus 58 16 616 718 2.2 15.2 1st fall Esopus 59 18 710 835 3.9 15.2 1st fall Esopus 60 10 726 852 4.6 15.2 1st fall Esopus 61 7 645 765 2.1 15.2 1st fall Esopus 62 12 620 725 2.5 15.2 1st fall Esopus 63 12 620 693 1.8 15.2 1st fall Esopus 64 13 627 740 2 15.2 1st fall Esopus 65 13 570 661 1.8 15.2 1st fall Esopus 66 6 642 739 2 15.2 1st fall Esopus 67 14 516 615 1.1 15.2 1st fall Esopus 68 13 600 704 1.8 15.2 1st fall Esopus 69 10 622 710 2 15.2 1st fall Esopus 70 10 589 681 1.7 15.2 1st fall Esopus 71 11 646 740 2.4 15.2 1st fall Esopus 72 8 681 777 2.5 15.2 1st fall Esopus 73 12 502 583 1.1 15.2 1st fall Esopus 74 13 570 672 1.6 15.2 1st fall Esopus 75 15 601 708 1.9 15.2 1st fall Esopus 76 17 705 811 2.8 15.2 1st fall Esopus 77 10 643 752 2.1 15.2 1st fall Esopus 78 11 682 783 2.8 15.2 1st fall Esopus 79 17 747 862 3 15.2 1st fall Esopus 80 13 703 816 3.4 15.2 1st fall Esopus 81 15 625 724 2 15.2 1st fall Esopus
95
82 12 662 774 2.5 15.2 1st fall Esopus 83 10 660 769 2.5 15.2 1st fall Esopus 84 14 617 715 2.1 15.2 1st fall Esopus 85 14 747 865 3.1 15.2 1st fall Esopus 86 18 726 826 3.4 15.2 1st fall Esopus 87 9 664 780 3.1 15.2 1st fall Esopus 88 12 626 730 2 15.2 1st fall Esopus 89 15 679 800 3 15.2 1st fall Esopus 90 9 675 786 2.2 15.2 1st fall Esopus 91 13 614 714 2.2 15.2 1st fall Esopus 92 - 753 872 4.2 15.2 1st fall Esopus 94 14 688 767 2.7 15.2 1st fall Esopus 95 11 660 760 2.1 15.2 1st fall Esopus 96 12 721 840 3.1 15.2 1st fall Esopus 97 17 620 708 2 15.2 1st fall Esopus 98 15 671 767 2.2 15.2 1st fall Esopus 99 10 722 839 3.1 15.2 1st fall Esopus 100 15 667 779 2.6 15.2 1st fall Esopus 101 13 741 853 3.3 15.2 1st fall Esopus 102 12 656 783 2.6 15.2 1st fall Esopus 103 14 743 841 3.1 15.2 1st fall Esopus 104 12 701 812 2.7 15.2 1st fall Esopus 105 12 642 743 2.1 15.2 1st fall Esopus 106 11 684 782 3.1 15.2 1st fall Esopus 107 13 630 740 2.2 15.2 1st fall Esopus 108 11 668 783 2.7 15.2 1st fall Esopus 109 10 640 730 1.9 15.2 1st fall Esopus 110 12 618 719 2.1 15.2 1st fall Esopus 111 11 633 723 2.5 15.2 1st fall Esopus 112 9 600 685 1.7 15.2 1st fall Esopus 113 13 632 733 2 15.2 1st fall Esopus 114 11 730 835 4.2 15.2 1st fall Esopus 115 - 726 831 3.7 15.2 1st fall Esopus 116 16 680 782 2.6 15.2 1st fall Esopus 117 11 719 820 2.5 15.2 1st fall Esopus 118 10 600 689 2.3 15.2 1st fall Esopus 119 8 621 701 2.1 15.2 1st fall Esopus 120 10 687 816 2.2 15.2 1st fall Esopus 121 14 605 692 2 15.2 1st fall Esopus 122 12 661 768 2.5 15.2 1st fall Esopus 123 14 669 761 2.6 15.2 1st fall Esopus 124 12 712 820 2.8 15.2 1st fall Esopus 125 11 693 792 2.7 15.2 1st fall Esopus 126 12 609 705 1.8 15.2 1st fall Esopus 127 12 642 741 2.2 15.2 1st fall Esopus 128 15 610 695 2.1 15.2 1st fall Esopus
96
129 13 657 752 2.4 15.2 1st fall Esopus 130 9 691 804 3.2 15.2 1st fall Esopus 131 13 648 741 2.5 15.2 1st fall Esopus 132 8 623 725 2.4 15.2 1st fall Esopus 133 14 574 665 1.9 17.8 1st fall Esopus 134 13 752 860 3.9 17.8 1st fall Esopus 135 - 832 943 4 17.8 1st fall Esopus 136 16 651 752 2.7 17.8 1st fall Esopus 137 12 761 880 4.6 17.8 1st fall Esopus 138 13 644 745 1.7 17.8 1st fall Esopus 139 12 784 907 4.8 17.8 1st fall Esopus 140 13 727 834 3.3 17.8 1st fall Esopus 141 11 647 762 3 17.8 1st fall Esopus 142 14 685 766 2.4 17.8 1st fall Esopus 143 16 790 916 3.4 17.8 1st fall Esopus 144 15 868 1004 5.5 17.8 1st fall Esopus 145 9 844 965 4.7 17.8 1st fall Esopus 146 15 710 821 3.1 17.8 1st fall Esopus 147 14 856 987 4.5 17.8 1st fall Esopus 148 11 866 801 2.1 17.8 1st fall Esopus 149 14 745 845 3.6 17.8 1st fall Esopus 150 14 776 907 5.1 17.8 1st fall Esopus 151 14 699 818 3.4 17.8 1st fall Esopus 152 11 783 918 4.4 17.8 1st fall Esopus 153 14 650 768 2.8 17.8 1st fall Esopus 154 11 719 832 4.7 17.8 1st fall Esopus 155 13 628 730 2.3 17.8 1st fall Esopus 156 12 711 803 2.9 17.8 1st fall Esopus 157 13 635 734 2.2 17.8 1st fall Esopus 158 15 753 864 4 17.8 1st fall Esopus 159 15 665 785 2.5 17.8 1st fall Esopus 160 12 705 830 3.1 17.8 1st fall Esopus 161 10 748 876 3.6 17.8 1st fall Esopus 162 12 652 737 2.3 17.8 1st fall Esopus 163 17 644 743 2.4 17.8 1st fall Esopus 164 14 703 803 2.9 17.8 1st fall Esopus 165 15 725 836 3.7 17.8 1st fall Esopus 166 13 760 881 3.5 17.8 1st fall Esopus 167 14 726 847 3.3 17.8 1st fall Esopus 168 10 680 776 3 17.8 1st fall Esopus 169 - 620 687 3.1 17.8 1st fall Esopus 170 16 795 936 4.4 17.8 1st fall Esopus 171 15 669 766 2.8 17.8 1st fall Esopus 172 12 780 921 4.3 17.8 1st fall Esopus 173 12 668 779 3 17.8 1st fall Esopus 174 14 680 805 3.2 15.2 1st fall Esopus
97
175 11 676 789 2.1 15.2 1st fall Esopus 176 19 681 801 2.6 15.2 1st fall Esopus 177 12 881 1001 4.4 15.2 1st fall Esopus 178 17 595 699 1.9 15.2 1st fall Esopus 179 12 780 896 4.5 15.2 1st fall Esopus 180 12 665 762 2.5 15.2 1st fall Esopus 181 17 655 782 2.2 15.2 1st fall Esopus 182 11 640 725 2.1 15.2 1st fall Esopus 183 15 600 712 2.3 15.2 1st fall Esopus 184 18 609 694 2.1 15.2 1st fall Esopus 185 16 688 793 3.2 15.2 1st fall Esopus 186 10 725 835 3.7 15.2 1st fall Esopus 187 14 621 723 2.1 15.2 1st fall Esopus 188 12 665 773 2.5 15.2 1st fall Esopus 189 8 636 742 2.1 15.2 1st fall Esopus 190 18 543 622 1.4 15.2 1st fall Esopus 191 13 630 732 2.2 15.2 1st fall Esopus 192 14 714 839 3.5 15.2 1st fall Esopus 193 - 665 784 2.45 15.2 1st fall Esopus
- - 630 722 1.8 15.2 1st fall Esopus - - 572 651 1.8 15.2 1st fall Esopus - - 612 718 - 15.2 1st fall Esopus - - 645 738 - 15.2 1st fall Esopus - - 557 656 - 15.2 1st fall Esopus - - 729 833 - 15.2 1st fall Esopus - - 638 729 - 15.2 1st fall Esopus - - 651 751 - 15.2 1st fall Esopus - 16 670 762 2.4 15.2 2nd spring Esopus
194 16 702 805 3.1 17.8 2nd spring Esopus 195 13 716 831 3.7 17.8 2nd spring Esopus 196 - 655 764 3 17.8 2nd spring Esopus 197 - 572 665 1.6 17.8 2nd spring Esopus 198 18 820 938 4 17.8 2nd spring Esopus 199 18 735 842 3.5 17.8 2nd spring Esopus 200 22 760 872 4.8 17.8 2nd spring Esopus 201 15 711 822 3.5 17.8 2nd spring Esopus 202 13 704 806 2.8 17.8 2nd spring Esopus 203 16 782 907 3.7 17.8 2nd spring Esopus 204 13 620 713 2.2 17.8 2nd spring Esopus 205 16 806 918 4.3 17.8 2nd spring Esopus 206 - 618 705 1.8 17.8 2nd spring Esopus 207 12 664 782 3 17.8 2nd spring Esopus 208 12 618 705 1.8 17.8 2nd spring Esopus 209 12 676 810 2.4 17.8 2nd spring Esopus 210 16 720 851 3.4 17.8 2nd spring Esopus 211 13 642 750 2.8 17.8 2nd spring Esopus
98
212 18 862 1000 6.1 17.8 2nd spring Esopus 213 12 738 845 4 17.8 2nd spring Esopus 214 14 761 900 4.1 17.8 2nd spring Esopus 215 12 636 732 2.4 17.8 2nd spring Esopus 216 14 675 785 3.2 17.8 2nd spring Esopus 217 16 907 1038 5.6 17.8 2nd spring Esopus 218 12 674 770 2.6 17.8 2nd spring Esopus 219 15 818 937 5.5 17.8 2nd spring Esopus 220 11 717 829 4 17.8 2nd spring Esopus 221 13 703 820 2.8 17.8 2nd spring Esopus 222 18 786 911 3.9 17.8 2nd spring Esopus 223 21 859 1004 5.3 17.8 2nd spring Esopus 224 17 562 652 2 17.8 2nd spring Esopus 225 - 774 893 3.6 17.8 2nd spring Esopus 226 16 776 897 4.5 17.8 2nd spring Esopus 227 15 755 893 5 17.8 2nd spring Esopus 228 26 860 984 4.7 17.8 2nd spring Esopus 229 14 685 802 3.1 17.8 2nd spring Esopus 230 11 638 760 2.3 17.8 2nd spring Esopus 231 16 685 810 2.9 17.8 2nd spring Esopus 232 11 667 791 2.2 17.8 2nd spring Esopus 233 11 591 675 1.7 17.8 2nd spring Esopus 234 16 660 778 2.2 17.8 2nd spring Esopus 235 10 626 706 2 17.8 2nd spring Esopus 236 12 587 689 1.9 17.8 2nd spring Esopus 237 12 656 748 2.3 15.2 2nd spring Esopus 238 15 779 895 4.7 15.2 2nd spring Esopus 239 16 686 805 2.8 15.2 2nd spring Esopus 240 15 658 750 2 15.2 2nd spring Esopus 241 11 647 760 2.9 15.2 2nd spring Esopus 242 17 604 684 2.3 15.2 2nd spring Esopus 243 13 685 795 3.2 15.2 2nd spring Esopus 244 10 646 747 2.2 15.2 2nd spring Esopus 245 9 596 699 1.7 15.2 2nd spring Esopus 246 16 732 853 3.1 15.2 2nd spring Esopus 247 14 688 804 2.1 15.2 2nd spring Esopus 248 13 653 740 2.3 15.2 2nd spring Esopus 249 11 605 703 2 15.2 2nd spring Esopus 250 12 698 801 3.1 15.2 2nd spring Esopus 251 10 703 811 2.5 15.2 2nd spring Esopus 252 12 659 775 2.6 15.2 2nd spring Esopus 253 16 760 882 3.4 15.2 2nd spring Esopus 254 12 666 749 2.8 15.2 2nd spring Esopus 255 - 767 811 3.4 15.2 2nd spring Esopus 256 13 650 770 2.1 15.2 2nd spring Esopus 257 14 596 690 2 15.2 2nd spring Esopus
99
258 13 664 788 2.5 15.2 2nd spring Esopus 259 14 712 830 3.4 15.2 2nd spring Esopus 260 11 644 759 2.5 15.2 2nd spring Esopus 261 15 828 916 4.8 15.2 2nd spring Esopus 262 10 656 781 2.1 15.2 2nd spring Esopus 263 11 670 775 2.8 15.2 2nd spring Esopus 264 13 727 845 3.2 15.2 2nd spring Esopus 265 11 625 733 1.8 15.2 2nd spring Esopus 266 12 689 798 2.8 15.2 2nd spring Esopus 267 12 660 766 2.9 15.2 2nd spring Esopus 268 11 651 749 2.7 15.2 2nd spring Esopus 269 9 544 636 1.2 15.2 2nd spring Esopus 270 13 735 833 4.6 15.2 2nd spring Esopus 271 18 668 775 2.7 15.2 2nd spring Esopus 272 - 751 850 3.1 15.2 2nd spring Esopus 273 10 705 803 2.6 15.2 2nd spring Esopus 274 13 700 817 2.8 15.2 2nd spring Esopus 275 17 639 749 2.4 15.2 2nd spring Esopus 276 14 722 837 3.4 15.2 2nd spring Esopus 277 17 754 880 4.3 15.2 2nd spring Esopus 278 10 655 778 2.3 15.2 2nd spring Esopus 279 11 706 805 3.2 15.2 2nd spring Esopus 280 19 682 784 2.4 15.2 2nd spring Esopus 281 13 658 769 2.6 15.2 2nd spring Esopus 282 16 660 763 2.6 15.2 2nd spring Esopus 283 13 623 732 2.3 15.2 2nd spring Esopus 284 12 673 797 2.8 15.2 2nd spring Esopus 285 18 715 841 3.3 15.2 2nd spring Esopus 286 11 629 726 2.4 15.2 2nd spring Esopus 287 9 690 779 2.7 15.2 2nd spring Esopus 288 16 670 786 2.8 15.2 2nd spring Esopus 289 - 702 815 2.5 15.2 2nd spring Esopus 290 14 664 773 2.6 15.2 2nd spring Esopus 291 10 621 711 2.7 15.2 2nd spring Esopus 292 12 685 798 3 15.2 2nd spring Esopus 293 12 625 721 2.6 15.2 2nd spring Esopus 294 17 740 843 3.7 15.2 2nd spring Esopus 295 13 674 787 3 15.2 2nd spring Esopus 296 14 618 715 1.9 10.2 2nd spring Esopus 297 - 571 666 1.2 10.2 2nd spring Esopus 298 13 779 892 3.8 10.2 2nd spring Esopus *299 8 476 552 0.7 10.2 2nd spring Esopus 300 8 483 571 0.9 10.2 2nd spring Esopus 301 14 595 685 1.9 10.2 2nd spring Esopus 302 14 728 830 3.3 10.2 2nd spring Esopus 303 5 477 562 0.7 10.2 2nd spring Esopus
100
304 11 639 718 1.7 10.2 2nd spring Esopus 305 12 694 794 2.6 10.2 2nd spring Esopus 306 12 665 745 2 10.2 2nd spring Esopus 307 7 480 566 0.8 10.2 2nd spring Esopus 308 14 630 726 1.8 10.2 2nd spring Esopus 309 13 648 745 2.2 10.2 2nd spring Esopus 310 12 632 749 2 10.2 2nd spring Esopus 311 6 472 549 0.8 10.2 2nd spring Esopus 312 10 580 680 1.5 10.2 2nd spring Esopus 313 5 500 596 0.8 10.2 2nd spring Esopus 314 5 468 555 0.7 10.2 2nd spring Esopus 315 15 644 749 2.2 10.2 2nd spring Esopus 316 13 557 638 1.3 10.2 2nd spring Esopus 317 9 534 620 1.1 10.2 2nd spring Esopus 318 13 629 735 2.2 10.2 2nd spring Esopus 319 7 508 595 1 10.2 2nd spring Esopus 320 5 419 490 0.5 10.2 2nd spring Esopus 321 5 441 526 0.5 10.2 2nd spring Esopus 322 12 670 783 2.2 10.2 2nd spring Esopus 323 5 485 562 0.8 10.2 2nd spring Esopus 324 11 568 660 1.6 10.2 2nd spring Esopus 325 6 448 630 0.6 10.2 2nd spring Esopus 326 8 554 655 1.3 10.2 2nd spring Esopus 327 8 462 648 1.2 10.2 2nd spring Esopus 328 15 793 915 3.9 10.2 2nd spring Esopus 329 11 568 682 1.7 10.2 2nd spring Esopus 330 9 622 740 1.9 10.2 2nd spring Esopus 331 12 696 796 2.8 10.2 2nd spring Esopus 332 9 583 671 1.5 10.2 2nd spring Esopus 333 9 542 630 1.2 10.2 2nd spring Esopus 334 6 480 559 0.7 10.2 2nd spring Esopus 335 13 - 846 3.5 15.2 3rd spring Albany 336 13 - 810 2.9 15.2 3rd spring Albany 337 13 685 805 2.5 15.2 3rd spring Albany 338 13 610 709 1.9 15.2 3rd spring Albany 339 - 761 883 3.2 15.2 3rd spring Albany 340 14 649 746 2.8 15.2 3rd spring Albany 341 14 681 791 3 15.2 3rd spring Albany 342 16 731 860 3.2 15.2 3rd spring Albany 343 13 737 870 3.6 15.2 3rd spring Albany 344 15 641 737 3 15.2 3rd spring Albany 345 16 705 800 3.5 15.2 3rd spring Albany 346 15 711 830 4.1 15.2 3rd spring Albany 347 14 745 870 3.1 15.2 3rd spring Albany 348 18 716 840 4.1 15.2 3rd spring Albany 349 17 665 751 3 15.2 3rd spring Albany
101
350 11 631 749 2.1 15.2 3rd spring Albany 351 25 758 865 4.1 15.2 3rd spring Albany 352 15 658 760 2.8 15.2 3rd spring Albany 353 14 690 800 2.5 15.2 3rd spring Albany 354 17 675 805 3 15.2 3rd spring Albany 355 14 652 742 1.9 15.2 3rd spring Albany 356 12 595 705 1.6 15.2 3rd spring Albany 357 13 731 841 3.5 15.2 3rd spring Albany 358 13 770 870 4 15.2 3rd spring Albany 359 13 726 850 3.7 15.2 3rd spring Albany 360 15 690 799 2.9 15.2 3rd spring Albany 361 10 655 751 2.3 15.2 3rd spring Albany 362 - 690 780 2.2 15.2 3rd spring Albany 363 11 625 710 1.9 15.2 3rd spring Albany 364 12 610 707 2.1 15.2 3rd spring Albany 365 14 691 803 2.7 15.2 3rd spring Albany 366 12 665 806 3.1 15.2 3rd spring Albany 367 10 706 805 2.9 15.2 3rd spring Albany 368 13 710 820 2.7 15.2 3rd spring Albany 369 14 735 850 4 15.2 3rd spring Albany 370 13 719 830 2.5 15.2 3rd spring Albany 371 16 634 769 3 15.2 3rd spring Albany 372 12 660 741 2.2 15.2 3rd spring Albany 373 13 681 800 2.9 15.2 3rd spring Albany 374 18 807 928 4.2 15.2 3rd spring Albany 375 16 678 805 2.5 15.2 3rd spring Albany 376 13 690 795 2.3 15.2 3rd spring Albany 377 12 695 790 2.3 15.2 3rd spring Albany 378 8 538 626 1.5 15.2 3rd spring Albany 379 16 765 880 4.2 15.2 3rd spring Albany 380 13 709 816 3 15.2 3rd spring Albany 381 15 670 775 2.3 15.2 3rd spring Albany 382 13 694 800 2.7 15.2 3rd spring Albany 383 14 679 775 3 15.2 3rd spring Albany 384 12 663 762 2.6 15.2 3rd spring Albany 385 18 758 900 3.3 15.2 3rd spring Albany 386 16 650 745 3 15.2 3rd spring Albany 387 15 718 807 3.2 15.2 3rd spring Albany 388 20 652 763 3.1 15.2 3rd spring Albany 389 - 660 758 2.5 15.2 3rd spring Albany 390 13 716 850 3.3 17.8 3rd spring Albany 391 - 791 910 3.4 17.8 3rd spring Albany 392 13 690 810 2.9 17.8 3rd spring Albany 393 18 774 890 5.1 17.8 3rd spring Albany 394 11 722 855 3.2 17.8 3rd spring Albany 395 14 640 746 2.2 17.8 3rd spring Albany
102
396 17 864 972 5.9 17.8 3rd spring Albany 397 13 820 919 3.5 17.8 3rd spring Albany 398 14 673 760 3.1 17.8 3rd spring Albany 399 13 752 885 4 17.8 3rd spring Albany 400 12 741 831 2.9 17.8 3rd spring Albany 401 - 686 799 2.3 17.8 3rd spring Albany 402 15 691 810 3.4 17.8 3rd spring Albany 403 17 731 832 3.9 17.8 3rd spring Albany 404 20 819 902 4 17.8 3rd spring Albany 405 15 702 812 3.2 17.8 3rd spring Albany 406 13 670 780 2.4 17.8 3rd spring Albany 407 14 680 780 2.4 17.8 3rd spring Albany 408 16 852 985 4.5 15.2 3rd spring Albany 409 - 779 910 3.9 15.2 3rd spring Albany 410 13 727 856 4 15.2 3rd spring Albany 411 14 711 811 2.8 15.2 3rd spring Albany 412 8 685 776 1.6 15.2 3rd spring Albany 413 14 760 875 3.7 15.2 3rd spring Albany 414 30 870 981 5.4 15.2 3rd spring Albany 415 12 657 751 2 15.2 3rd spring Albany 416 16 793 930 4 15.2 3rd spring Albany 417 19 833 950 6.2 15.2 3rd spring Albany 418 14 636 761 2.1 15.2 3rd spring Albany 419 11 590 684 1.3 15.2 3rd spring Albany 420 14 771 884 3.3 15.2 3rd spring Albany 421 13 714 811 2.5 15.2 3rd spring Albany 422 19 824 941 4.7 15.2 3rd spring Albany 423 17 759 871 3 15.2 3rd spring Albany 424 13 630 847 3.7 15.2 3rd spring Albany 425 17 700 823 2.5 15.2 3rd spring Albany 426 14 809 947 4.7 15.2 3rd spring Albany 427 15 650 750 2.1 15.2 3rd spring Albany 428 15 683 790 2.9 15.2 3rd spring Albany 429 29 744 830 2.5 15.2 3rd spring Albany 430 13 872 1010 4.5 15.2 3rd spring Albany 431 14 655 750 2.7 15.2 3rd spring Albany 432 15 755 860 4.2 15.2 3rd spring Albany 433 16 675 789 2.1 15.2 3rd spring Albany 434 14 742 865 3 15.2 3rd spring Albany 435 16 790 872 4.6 15.2 3rd spring Albany 436 16 780 901 4 15.2 3rd spring Albany 437 14 836 946 5.3 15.2 3rd spring Albany 438 13 668 789 3.1 15.2 3rd spring Albany 439 15 651 763 2.4 15.2 3rd spring Albany 440 15 712 760 3.3 15.2 3rd spring Albany 441 12 762 886 3.6 15.2 3rd spring Albany
103
442 15 685 800 3 15.2 3rd spring Albany 443 18 822 940 4.9 15.2 3rd spring Albany 444 15 725 835 4 15.2 3rd spring Albany 445 17 750 865 3.6 15.2 3rd spring Albany 446 - 802 942 4.5 15.2 3rd spring Albany 447 10 577 662 1.5 10.2 3rd spring Albany 448 12 684 801 2.6 10.2 3rd spring Albany 449 11 661 760 2.1 10.2 3rd spring Albany 450 11 641 755 1.7 10.2 3rd spring Albany 451 9 669 777 1.9 10.2 3rd spring Albany 452 14 697 805 2.6 10.2 3rd spring Albany 453 14 596 685 1.4 10.2 3rd spring Albany 454 15 667 780 3.7 10.2 3rd spring Albany 455 12 574 660 1.5 10.2 3rd spring Albany 456 7 550 632 1.5 10.2 3rd spring Albany 457 6 575 660 1.7 10.2 3rd spring Albany 458 16 724 838 3 10.2 3rd spring Albany 459 9 573 649 1.5 10.2 3rd spring Albany 460 14 632 741 1.9 10.2 3rd spring Albany 461 14 733 865 3.7 10.2 3rd spring Albany 462 - 520 590 0.9 10.2 3rd spring Albany 463 15 675 800 3.1 10.2 3rd spring Albany 464 5 450 530 0.8 10.2 4th summer Burden 465 16 745 860 4.8 15.2 4th summer Burden 466 13 610 693 2.3 15.2 4th summer Burden 467 8 568 686 1.5 15.2 4th summer Burden 468 15 725 831 2.6 15.2 4th summer Burden 469 - 745 850 3.6 15.2 4th summer Burden 470 17 775 895 4.1 15.2 4th summer Burden 471 - 750 875 3.3 15.2 4th summer Burden 472 13 710 805 2.8 15.2 4th summer Burden 473 15 655 772 2.2 15.2 4th summer Burden 474 12 695 807 2.8 15.2 4th summer Burden 475 18 705 815 2.7 15.2 4th summer Burden 476 11 605 695 1.7 15.2 4th summer Burden 477 10 725 850 3.8 15.2 4th summer Burden 478 13 700 805 2.8 15.2 4th summer Burden 479 13 735 835 3.9 15.2 4th summer Burden 480 14 715 805 3.5 15.2 4th summer Burden 481 13 815 945 4.8 15.2 4th summer Burden 482 15 705 800 2.7 15.2 4th summer Burden 483 16 705 815 3.5 15.2 4th summer Burden 484 14 695 795 2.4 15.2 4th summer Burden 485 15 825 945 5 15.2 4th summer Burden 486 18 750 880 4.3 15.2 4th summer Burden 487 16 750 880 3.8 15.2 4th summer Burden
104
488 16 665 770 2 15.2 4th summer Burden 489 15 655 770 2.4 15.2 4th summer Burden 490 11 580 683 1.4 15.2 4th summer Burden 491 16 675 760 2.5 15.2 4th summer Burden 492 16 740 865 3.4 15.2 4th summer Burden 493 12 710 795 2.5 15.2 4th summer Burden 494 11 700 810 2.7 15.2 4th summer Burden 495 5 580 665 1.2 15.2 4th summer Burden 496 9 526 605 1 15.2 4th summer Burden 497 16 655 745 2.4 15.2 4th summer Burden 498 17 475 775 3.3 15.2 4th summer Burden 499 - 640 725 2.2 15.2 4th summer Burden 500 14 665 770 2.1 15.2 4th summer Burden 501 11 705 805 2.8 15.2 4th summer Burden 502 12 610 712 1.7 15.2 4th summer Burden 503 13 660 755 3.4 15.2 4th summer Burden 504 10 705 815 2.9 15.2 4th summer Burden 505 20 770 905 5 15.2 4th summer Burden 506 14 755 890 3.8 15.2 4th summer Burden 507 12 680 760 2.5 15.2 4th summer Burden 508 14 670 780 2.9 15.2 4th summer Burden 509 16 675 755 2.8 15.2 4th summer Burden 510 12 650 755 2.3 15.2 4th summer Burden 511 14 695 785 3.8 15.2 5th summer Burden 512 23 925 1045 9 15.2 5th summer Burden 513 17 710 805 3.2 15.2 5th summer Burden 514 13 735 845 3.2 15.2 5th summer Burden 515 15 720 835 2.8 15.2 5th summer Burden 516 16 695 785 2.6 15.2 5th summer Burden 517 13 775 895 4.5 15.2 5th summer Burden 518 17 745 845 4.2 15.2 5th summer Burden 519 15 735 855 3.8 15.2 5th summer Burden 520 15 690 790 2.9 15.2 5th summer Burden 521 16 670 795 2.8 15.2 5th summer Burden 522 12 695 795 2.5 15.2 5th summer Burden 523 14 685 780 3 15.2 5th summer Burden 524 13 700 815 2.9 15.2 5th summer Burden 525 16 760 860 4.2 15.2 5th summer Burden 526 10 645 730 2.1 15.2 5th summer Burden 527 16 695 805 2.7 15.2 5th summer Burden 528 13 680 785 3.2 15.2 5th summer Burden 529 8 635 745 2.2 15.2 5th summer Burden 530 13 720 840 2 15.2 5th summer Burden 531 14 715 825 2 15.2 5th summer Burden 532 13 690 805 2.6 15.2 5th summer Burden 533 14 820 955 3.7 15.2 5th summer Burden
105
534 17 790 925 4 15.2 5th summer Burden 535 18 790 920 4.4 15.2 5th summer Burden 536 14 695 820 2 15.2 5th summer Burden 537 13 640 735 2.5 15.2 5th summer Burden 538 16 745 840 3.4 15.2 5th summer Burden 539 19 845 955 5.5 15.2 5th summer Burden 540 12 730 845 2.6 15.2 5th summer Burden 541 13 760 895 3 15.2 5th summer Burden 542 16 765 865 2.8 15.2 5th summer Burden 543 15 610 725 2 15.2 5th summer Burden 544 22 690 795 2 15.2 5th summer Burden 545 14 860 990 5.2 15.2 5th summer Burden 546 13 635 705 1.1 15.2 5th summer Burden 547 - 595 660 1.5 10.2 6th fall Esopus 548 13 670 763 2.8 10.2 6th fall Esopus 549 - 675 740 2.3 10.2 6th fall Esopus 550 13 695 771 2.25 10.2 6th fall Esopus 551 7 520 558 1 10.2 6th fall Esopus 552 - 563 640 1.55 10.2 6th fall Esopus 553 12 648 707 2.15 10.2 6th fall Esopus 554 14 755 834 3.75 10.2 6th fall Esopus 555 13 647 758 2.7 10.2 6th fall Esopus 556 8 505 592 1.15 10.2 6th fall Esopus 557 11 580 680 2 10.2 6th fall Esopus 558 10 673 790 2.6 15.2 6th fall Esopus 559 23 704 780 3.2 15.2 6th fall Esopus 560 13 715 832 3.05 15.2 6th fall Esopus 561 12 710 805 3.3 15.2 6th fall Esopus 562 14 615 710 2.45 15.2 6th fall Esopus 563 - 660 768 2.55 15.2 6th fall Esopus 564 10 668 765 2.3 15.2 6th fall Esopus 565 13 570 680 1.95 15.2 6th fall Esopus 566 13 660 770 2.1 15.2 6th fall Esopus 567 22 760 885 3.8 15.2 6th fall Esopus 568 14 665 770 2.85 15.2 6th fall Esopus 569 11 640 730 2.2 15.2 6th fall Esopus 570 17 790 940 4.4 15.2 6th fall Esopus 571 15 700 834 2.7 15.2 6th fall Esopus 572 12 636 720 2.1 15.2 6th fall Esopus 573 15 705 804 3.05 15.2 6th fall Esopus 574 12 640 724 2.2 15.2 6th fall Esopus 575 13 588 663 1.65 15.2 6th fall Esopus 576 13 645 750 2.3 15.2 6th fall Esopus 577 14 700 760 2.7 15.2 6th fall Esopus 578 11 700 812 2.75 15.2 6th fall Esopus 579 8 495 583 1.05 15.2 6th fall Esopus
106
580 15 740 850 3.45 15.2 6th fall Esopus 581 15 680 780 2.6 15.2 6th fall Esopus 582 13 680 775 2.55 15.2 6th fall Esopus 583 13 450 630 1.35 15.2 6th fall Esopus
- - 575 665 1.6 15.2 6th fall Esopus - - 620 730 2.4 15.2 6th fall Esopus - - 660 740 2.4 15.2 6th fall Esopus - - 675 770 3.55 15.2 6th fall Esopus - - 695 778 3.3 15.2 6th fall Esopus - - 685 790 2.7 15.2 6th fall Esopus
.
107
References
Anderson, M. R., and R. M. Neumann. 2000. Indirect estimate of gill net mesh size selectivity for landlocked alewives. Journal of Freshwater Ecology 15(4):491-496.
Bain, M. B. 1997. Atlantic and shortnose sturgeons of the Hudson River: common
and divergent life history attributes. Environmental Biology of Fishes 48:347-358.
Bain, M. B., S. Nack, W. Dovel. 1997. Trends in the abundance of Hudson River sturgeons. In Sturgeon Notes #4. Department of Natural Resources, Cornell University. (previously available at http://www.dnr.cornell.edu/hydro2/sturgeon/sn4/sn4_1-17.htm#senescence)
Bain, M. B., N. Haley, D. L. Peterson, K. K. Arend, K. E. Mills, and P. J. Sullivan. 2000. Shortnose sturgeon in the Hudson River: an Endangered Species recovery story. Biology, Management and Protection of Sturgeon Symposium Pre-Print. EPRI, Palo Alto, CA.
Baird, D., R. R. Christian, C. H. Peterson, and G. A. Johnson. 2004. Consequences of
hypoxia on estuarine ecosystem function: energy diversion from consumers to microbes. Ecological Applications 14:805-822.
Baranov, F.I. 1914. The capture of fish by gillnets. Materialy Poznaniyu Russ.
Rybolov. 3(6):56-99 (as referenced by Hamley 1975). Beamish, R. J., and D. A. Fournier. 1981. A method for comparing the precision of a
set of age determination. Canadian Journal of Fisheries and Aquatic Sciences. 38:982-983.
Bemis, W. E., and B. Kynard. 1997. Sturgeon rivers: an introduction to
acipenseriform biogeography and life history. Environmental Biology of Fishes 48:167-183.
Bertalanffy, L. v. 1938. A quantitative theory on organic growth. Human Biology 10:181-213. Boreman, J. 1997. Sensitivity of North American sturgeons and paddlefish to fishing
mortality. Environmental Biology of Fishes 48:399-405. Boyle, R. H. 1979. The Hudson River: A natural and unnatural history. W.W. Norton
108
& Company. New York. Brennan, J. S., and G. M. Cailliet. 1989. Comparative age-determination techniques
for white sturgeon in California. Transactions of the American Fisheries Society 118:296-310.
Brennan, J. S., and G. M. Cailliet. 1991. Age determination and validation studies of white sturgeon, Acipenser transmontanus, in California. Pages 209-234 in P. Williot, editor. Acipenser. Centre National du Machinisme Agricole du Genie Rural des Eaux et des Forets, Bordeaux. Buckley, J., and B. Kynard. 1985. Habitat use and behavior of pre-spawning and spawning shortnose sturgeon, Acipenser brevirostrum, in the Connecticut
River. Pages 111-117 in F. P. Binkowski, and S. I. Doroshov, editors. North American Sturgeons: Biology and Aquaculture Potential. Dr. W. Junk Publishers, Dordercht.
Burton, E. J., A. H. Andrews, K. H. Coale, and G. M. Cailliet. 1999. Application of
Radiometric age determination to three long-lived fishes using 210Pb:226Ra disequilibria in calcified structures: A review. Pages 77-87 in J.A. Musick editor. Life in the Slow Lane: Ecology and Conservation of Long-lived Marine Animals. American Fisheries Society, Symposium 23, Bethesda, Maryland
Campana, S. E. 1984. Microstructural growth patterns in the otoliths of larval and juvenile starry flounder, Platichthys stellatus. Canadian Journal of Zoology
62:1507-1512. Campana, S. E., M. C. Annad, and J. I. McMillan. 1995. Graphical and statistical
methods for determining the consistency of age determinations. Transactions of the American Fisheries Society 124:131-138.
Campana, S. E. 2001. Accuracy, precision and quality control in age determination,
including a review of the use and abuse of age validation methods. Journal of Fish Biology 59:197-242.
Campbell, J. G., and L. R. Goodman. 2004. Acute sensitivity of juvenile shortnose sturgeon to low dissolved oxygen concentrations. Transactions of the
American Fisheries Society 133:772-776. Carlson, D. M., and K. W. Simpson. 1987. Gut contents of juvenile shortnose
sturgeon in the upper Hudson estuary. Copeia 3:796-802. Casselman, J. M. 1987. Determination of age and growth. Pages 209-242 in A. H.
109
Weatherley and H. S. Gill, editors. The Biology of Fish Growth. Academic Press, Inc. Orlando, FL.
Chang, W.Y.B. 1982. A statistical method for evaluating the reproducibility of age
determination. Canadian Journal of Fisheries and Aquatic Sciences 39:1208-1210.
Collins, M. R., and T. I. J. Smith. 1996. Sturgeon fin ray removal is nondeleterious. North American Journal of Fisheries Management 16:939-941. Collins, M. R., and T. I. J. Smith. 1997. Distributions of shortnose and Atlantic
sturgeons in South Carolina. North American Journal of Fisheries Management 17:995-1000.
Collins, M. R., S. G. Rogers, T. I. J. Smith, and M. L. Moser. 2000. Primary factors affecting sturgeon populations in the Southeastern United States: fishing
mortality and degradation of essential habitats. Bulletin of Marine Science 66(3):917-928.
Coutant, C. C. 1985. Striped bass, temperature, and dissolved oxygen: a speculative
hypothesis for environmental risk. Transactions of the American Fisheries Society 114: 31-61.
Crecco, V. A., and T. F. Savoy. 1984. Effects of fluctuations in hydrographic
conditions on year-class strength of American Shad (Alosa sapidissima) in the Connecticut River. Canadian Journal of Fisheries and Aquatic Science 41:1216-1223.
Cuerrier, J. P. 1951. The use of pectoral fin rays for determining age of sturgeon and
other species of fish. Canadian Fish Culturist 11:10-18. Dadswell, M. J. 1979. Biology and population characteristics of the shortnose
sturgeon, Acipenser brevirostrum LeSueur 1818 (Osteichythyes: Acipenseridae), in the Saint John River Estuary, New Brunswick, Canada. Canadian Journal of Zoology 57:2186-2210.
Dadswell, M. J., B. D. Taubert, T. S. Squiers, D. Marchette, and J. Buckley. 1984. Synopsis of Biological Data on Shortnose Sturgeon, Acipenser brevirostrum LeSueur 1818. NOAA Technical Report NMFS 14 and FAO Fisheries
Synopsis 140. De Vlaming, V. L. 1972. Environmental control of teleost reproductive cycles: a
brief review. Journal of Fish Biology. 4: 131-140. Doroshov, S. I., G. P. Moberg, and J. P. van Eenennaam. 1997. Observation on the
110
reproductive cycle of cultured white sturgeon, Acipenser transmontanus. Environmental Biology of Fishes 48:265-278.
Dovel, W. L. 1981. The endangered shortnose sturgeon of the Hudson estuary: Its
life history and vulnerability to the activities of man. The Oceanic Society. FERC Contract No. DE-AC 39-79 RC-10074.
Dovel, W. L., A. W. Pekovitch, and T. J. Berggren. 1992. Biology of the shortnose sturgeon (Acipenser brevirostrum LeSueur, 1818) in the Hudson River
Estuary, New York. Pages 187-216 in C. L. Smith, editor. Estuarine Research in the 1980s. State University of New York Press, Albany, NY.
Doyon, C., R. Fortin, and P. A. Spear. 1999. Retinoic acid hydroxylation and
teratogenesis in lake sturgeon (Acipenser fulvescens) from the St. Lawarence River and Abitibi region, Quebec. Canadian Journal of Fisheries and Aquatic Sciences 56:1428-1436.
Duncan, M. S., J. J. Isley, and D. W. Cooke. 2004. Evaluation of shortnose sturgeon
spawning in the Pinopolis Dam tailrace, South Carolina. North American Journal of Fisheries Management. 24: 932-938.
Eby, L. A., L. B. Crowder, C. M. McClellan, C. H. Peterson, and M. J. Powers.
Habitat degradation from intermittent hypoxia: impacts on demersal fishes. 2005. Marine Ecology Progress Series 291:249-261.
Fry, F. E. J. 1971. The effect of environmental factors on the physiology of fish.
Pages 1-98 in W. S. Hoar, and D. J. Randall, editors. Fish Physiology, Volume IV. Academic Press, New York.
Geoghegan, P., M. T. Mattson, and R. G. Keppel. 1992. Distribution of the shortnose sturgeon in the Hudson River Estuary, 1984-1988. Pages 217-227 in C. L.
Smith, editor. Estuarine Research in the 1980s. State University of New York Press, Albany, NY.
Gross, M. R., J. Repka, C. T. Robertson, D. H. Secor, and W. van Winkle. 2002. Sturgeon Conservation: Insights from Elasticity Analysis. American Fisheries Society Symposium 28:13-30. Guenette, S., and coauthors. 1992. La périodicité de la croissance chez la femelle de l'esturgeon jaune (Acipenser fulvescens) du fleuve Saint-Laurent est-elle reliée
a la périodicité de la reproduction? Canadian Journal of Fisheries and Aquatic Science 49:1336-1341. Greeley, J. R. 1937. XI. Fishes of the area with annotated list. Pages 45-103 in A
Biological Survey of the Lower Hudson Watershed. Supplement to 26th
111
Annual Report,1936. New York State Conservation Department, Albany, New York.
Gulland, J. A. 1983. Fish Stock Assessment: A Manual of Basic Methods. FAO
Wiley Series on Food and Agriculture, Rome.Volume 1. Haas, R. E., and C. W. Recksiek. 1995. Age verification of winter flounder in
Narragansett Bay. Transactions of the American Fisheries Society 124:103-111.
Haley, N. J. 1999. Habitat characteristics and resource use patterns of sympatric
sturgeons in the Hudson River estuary. Masters Thesis, University of Massachusetts. Amherst.
Hamley, J. M., and H. A. Regier. 1973. Direct estimates of gillnet selectivity to
walleye (Stizostedion vitreum vitreum). Journal of Fisheries Resources Board of Canada 30:817-830.
Hamley, J. M. 1975. Review of gillnet selectivity. Journal of Fisheries Resources
Board of Canada 32(11):1943-1969. Hansen, L. G., W. B. Wiekhorst, and L. Simon. 1976. Effects of dietary Aroclor 1242
on channel catfish (Ictalurus punctatus) and the selective accumulation of PCB components. Journal of Fisheries Resources Board of Canada 33:1343-1352.
Hilborn, R., and C. J. Walters. 2001. Quantitative Fisheries Stock Assessment:
Choice Dynamics and Uncertainty, 3rd edition. Kluwer Academic Publishers, Norwell, MA.
Holt, S. J. 1963. A method for determining gear selectivity and its application.
ICNAF Special Publication 5:106-115. Hoenig, J. M. 1983. Empirical use of longevity data to estimate mortality rates.
Fisheries Bulletin 82:898-903. Hovgard, H. 1996. A two-step approach to estimating selectivity and fishing power of research gillnets used in Greenland waters. Canadian Journal of Fisheries and Aquatic Science 53:1014-1117. Hovgard, H., and H. Lassen. 2000. Manual on estimation of selectivity for gillnet and longline gears in abundance surveys, volume 397. FAO, Rome. Hughes, N. F. 1998. Reduction in growth due to electrofishing and tagging may
112
change interannual movement behavior of stream salmonids: evidence from Arctic grayling in an interior Alaskan stream. Transactions of the American Fisheries Society 127:1072-1077.
Hurley, K. L., R. J. Sheehan, and R. C. Heidinger. 2004. Accuracy and precision of
age estimates for pallid sturgeon from pectoral fin rays. North American Journal of Fisheries Management 24:715-718.
Jenkins, W. E., T. I. J. Smith, L. D. Heyward, and D. M. Knott. 1993. Tolerance of
shortnose sturgeon, Acipenser brevirostrum, juveniles to different salinity and dissolved oxygen concentrations. Proceedings of the 47th Annual Conference of the Southeastern Association of Fish and Wildlife Agencies 47:476-484.
Jung, S., and E. D. Houde. 2003. Spatial and temporal variabilities of pelagic fish community structure and distribution in Chesapeake Bay, USA. Estuarine,
Coastal and Shelf Science 58:335-351. Kohlhorst, D. W., L. W. Miller, and J. J. Orsi. 1980. Age and growth of white
sturgeon collected in the Sacramento-San Joaquin Estuary, California: 1965-1970 and 1973-1976. California Fish and Game 66(2):83-95.
Kynard, B. 1997. Life history, latitiudinal patterns, and status of the shortnose
sturgeon, Acipenser brevirostrum. Environmental Biology of Fishes 48:319-334.
Kynard, B., and M. Horgan. 2002. Ontogenetic behavior and migration of Atlantic
stgurgeon, Acipenser oxyrinchus oxyrinchus, and shortnose sturgeon, A. brevirostrum, with notes on social behavior. Environmental Biology of Fishes 63:137-150.
Leslie, J. A., K. A. Abood, E. A. Maikish, and P. J. Keeser. 1988. Recent dissolved oxygen trends in the Hudson River. Pages 287-303 in C. L. Smith, editor. Fisheries Research in the Hudson River. State University of New York Press,
Albany. Lough, R. G., M. Pennington, G. R. Goltz, and A. A. Rosenberg. 1982. Age and
growth of larval Atlantic herring, Clupea harengus L., in the Gulf of Maine-Georges Bank region based on otolith gorwth increments. Fishery Bulletin 80(187-199).
Maartens, L., A. J. Booth, and T. Hecht. 1999. The growth of monkfish Lophius vomerinus in Nambian waters, with a comparison of otolith and illicia
methods of ageing. Fisheries Research 44:139-148. Malone, J. C., G. E. Forrester, and M. A. Steele. 1999. Effects of subcutaneous
113
microtags on the growth, survival, and vulnerability to predation of small reef fishes. Journal of Experimental Marine Biology and Ecology 237:243-253.
Maurice, K. R., R. W. Blye, P. L. Harmon, and D. Lake. 1987. Increased spawning by American shad coincident with improved dissolved oxygen in the tidal
Delaware River. American Fisheries Society Symposium (1):79-88. McCarthy, I. D., L. A. Fuiman, and M. C. Alvarez. 2003. Aroclor 1254 affects
growth and survival skills of Atlantic croaker Micropogonias undulatus larvae. Marine Ecology Progress Series 252:295-301.
Milton, D. A., D. Die, C. Tenakanai, and S. Swales. 1998. Selectivity for barramundi (Lates calcarifer) in the Fly River, Papua New Guinea: implications for
managing gill-net fisheries on protandrous fishes. Marine and Freshwater Research 49:499-506.
Moser, M. L., M. Bain, M. R. Collins, N. Haley, B. Kynard, J. C. O’Herron II, G.
Rogers,and T. S. Squiers. 2000. A protocol for use of shortnose and Atlantic sturgeons. NOAA Technical Memorandum NMFS-OPR-18:1-18.
Nack, S. Personal Communication. Columbia County Soil and Water Conservation
District. Ghent, New York. November 2004. Nakamoto, R. J., T. T. Kisanuki, and G. H. Goldsmith. 1995. Age and growth of
Klamath River green sturgeon (Acipenser medirostris). U.S. Fish and Wildlife Service, 93-FP-13, Yreka.
Ndayibagira, A., M. J. Cloutier, P. D. Anderson, and P. A. Spear. 1995. Effects of 3,3',4,4'-Tetrachlorobiphenyl on the dynamics of vitamin-A in brook trout (Salvelinus fontinalis) and intestinal retinoid concentrations in lake sturgeon (Acipenser fulvescens). Canadian Journal of Fisheries and Aquatic Sciences
52: 512-520. Niklitschek, E. J. 2001. Bioenergetics modeling and assessment of suitable habitat for
juvenile Atlantic and shortnose sturgeons (Acipenser oxyrinchus and A. brevirosturm) in the Chesapeake Bay. Doctoral dissertation. University of Maryland, College Park.
Niklitschek, E. J., and D. H. Secor. 2005. Modeling spatial and temporal variation of
suitable nursery habitats for Atlantic surgeon in the Chesapeake Bay. Estuarine, Coastal and Shelf Science 64:135-148.
Nilo, P., P. Dumont, and R. Fortin. 1997. Climatic and hydrological determinants of
114
year-classs strength of St. Lawrence River lake sturgeon (Acipenser brevirostrum). Canadian Journal of Fisheries and Aquatic Science 54:774-780.
NMFS (National Marine Fisheries Service). 1998. Recovery Plan for the Shortnose Sturgeon (Acipenser brevirostrum). Prepared by the Shortnose Sturgeon
Recovery Team, for the National Marine Fisheries Service Silver Spring, MD. NYSDEC (New York State Department of Environmental Conservation) 2001.
Injuries to Hudson River fishery resources: fishery closures and consumption restrictions, Hudson River natural resource damage assessment. NYSDEC Final report.
Ott, R. L., and M. Longnecker. 2001. An introduction to statistical methods and data
anlaysis, 5th edition. Duxbury-Thompson Learning, Pacific Grove, CA. Paragamian, V. L., and V. D. Wakkinen. 2002. Temporal distribution of Kooteni
River white sturgeon spawning events and the effect of flow and temperature. Journal of Applied Ichthyology 18:542-549.
Paragamian, V. L., and R. C. Beamesderfer. 2003. Growth estimates from tagged
white sturgeon suggest that ages from fin rays underestimate true age in the Kootenai River, USA and Canada. Transactions of the American Fisheries Society 132:895-903.
Parrish D. L., B. Vondracek, and W. J. Eckmayer. 1994. Accuracy and precision of
daily age estimates for walleyes from Ohio hatcheries and Lake Erie. Transactions of the American Fisheries Society 123:601-605.
Richmond, A. M., and B. Kynard. 1995. Ontogenetic behavior of shortnose sturgeon,
Acipenser brevirostrum. Copeia 1:172-182. Ricker, W. E. 1975. Computation and interpretation of biological statistics of fish
populations. Department of Environment Fisheries and Marine Service, Ottawa.
Rien, T. A., and R. C. Beamesderfer. 1994. Accuracy and precision of white sturgeon
age estimates from pectoral fin rays. Transactions of the American Fisheries Society 123:255-265.
Root, K. V. 2002. Evaluating risks for threatened aquatic species: the shortnose
sturgeon in the Connecticut River. American Fisheries Society Symposium 28:45-54.
Rossiter, A., D. L. G. Noakes, and F. W. H. Beamish. 1995. Validation of age
115
estimation for the lake sturgeon. Transactions of the American Fisheries Society 124:777-781.
Roussow, G. 1957. Some considerations concerning sturgeon spawning periodicity. Journal of the Fisheries Resource Board of Canada 14:553-572. Salvanes, A. G. V. 1991. The selectivity for cod (Gadus morhua L.) in two
experimental trammel-nets and one gillnet. Fisheries Research 10:265-285. Scheirer, J. W., and D. W. Coble. 1991. Effects of Floy FD-67 anchor tags on growth
and condition of northern pike. North American Journal of Fisheries Management 11:369-373.
Secor, D. H., and T. E. Gunderson. 1998. Effects of hypoxia and temperature on
survival, growth, and respiration of juvenile Atlantic strugeon, Acipenser oxyrinchus. Fishery Bulletin 96:603-613.
Secor, D. H., and J. R. Waldman. 1999. Historical abundance of Delaware Bay
Atlantic sturgeon and potential rate of recovery. American Fisheries Society Symposium 23:203-216.
Secor, D. H. 2000. Spawning in the nick of time? Effect of adult demographics on
spawning behaviour and recruitment in Chesapeake Bay striped bass. ICES Journal of Marine Sciences 157:403-411. Secor, D. H., and E. J. Niklitschek. 2001. Hypoxia and sturgeons. Chesapeake
Biological Laboratory, TS-314001-CBL, Solomons, MD. Secor, D. H., P. J. Anders, W. van Winkle, and D. A. Dixon. 2002. Can we study sturgeons to extinction? What we do and don't know about the conservation of North American sturgeons. American Fisheries Society Symposium 28:3-10. Sokolov, L. I., and N. V. Akimova. 1976. Age determination of the Lena River
sturgeon. Journal of Ichthyology 16(5):773-778. Stevens, D. E., H. K. Chadwick, and R. E. Painter. 1987. American shad and striped
bass in California's Sacramentro-San Joaquin River system. Pages 66-78 in. American Fisheries Society.
Stevenson, J. T. 1997. Life history characteristics of Atlantic sturgeon (Acipenser
oxyrinchus) in the Hudson River and a model for fishery management. Master’s thesis. University of Maryland, College Park.
Stevenson, J. T., and D. H. Secor. 1999. Age determination and growth of Hudson
River Atlantic sturgeon, Acipenser oxyrinchus. Fishery Bulletin 97:153-166.
116
Taubert, B. D. 1980. Reproduction of shortnose sturgeon (Acipenser brevirostrum) in Holyoke Pool, Connecticut River, Massachusetts. Copeia 1:114-117.
US Geological Survey (USGS). 2005. Monthly/annual flow data. Green Island, NY
monitoring station # 01358000. Available: http://nwis.waterdata.usgs.gov (May 2005).
Veinott, G. I., and R. D. Evans. 1999. An examination of elemental stability in the fin
ray of the white sturgeon with Laser Ablation Sampling-Inductively Coupled Plasma-Mass Spectrometry (LAS-ICP-MS). Transactions of the American Fisheries Society 128:352-361.
Veshchev, P. V. 1982. Reproduction of sterlet, Acipenser ruthenus (Acipenseridae),
in the Lower Volga. Journal of Ichthyology 22(4):40-47. Vladykov, V. D., and J. R. Greeley. 1963. Order Acipenseroidei. Pages 24-60 in V.
H. Olsen, editor. Fishes of the Western North Atlantic, Part III. Memoirs of the Sears Foundation for Marine Research, New Haven, CT.
Warner, R. R., and P. L. Chesson. 1985. Coexistence mediated by recruitment
fluctuations: a field guide to the storage effect. The American Naturalist 125:769-787.
Webb, M. A. H., J. P. van Eenennaam, S. I. Doroshov, and G. P. Moberg. 1999. Preliminary observations on the effects of holding temperature on
reproductive performance of female white sturgeon, Acipenser transmontanus Richardson. Aquaculture 176:315-329.
Webb, M. A. H., J. P. van Eenennaam, G. W. Feist, J. Linares-Casenave, M. S.
Fitzpatrick, C. B. Schreck, and S. I. Doroshov. 2001. Effects of thermal regime on ovarian maturation and plasma sex steroids in farmed white surgeon, Acipenser transmontanus. Aquaculture 201:137-151.
Weatherley, A. H., and H. S. Gill. 1987. The Biology of Fish Growth. Academic
Press, Inc. Orlando, FL. Winnemiller, K. O., and K. A. Rose. 1992. Patterns of life-history diversification in
North American fishes: implication for population regulation. Canadian Journal of Fisheries and Aquatic Sciences 49:2,196-2,218.
Whiteman, K. W., V. H. Travnichek, M. L. Wildhaber, A. DeLonay, D. Papoulias,
and D. Tillett.. 2004. Age estimation for shovelnose sturgeon: a cautionary note based on annulus formation in pectoral fin rays. North American Journal of Fisheries Management 24:731-734.
117
Zar, J. H. 1999. Biostatistical Analysis 4th edition. Pearson Education, Inc. Dehli,
India.